The MRC guidance emphasizes the incorporation of both theory and best available evidence to develop complex interventions [34], while the BCW provides a systematic process through which to develop such interventions by completing a series of activities in stages [32]. This research was guided by stage 1 (understand the behavior) and stage 3 (identify content and implementation options) of the BCW. Stage 2 (identify intervention options) was not conducted, as an appropriate behavior change strategy was identified in phase 1 (stage 1 of the BCW guidance), which was directly mapped to behavior change techniques (BCTs) in phase 2 (stage 3 of the BCW guidance). In phase 3, the intervention and implementation strategies were refined through engagement with patient and public involvement (PPI) group members, cycling instructors, and experts in the field of behavior change. Figure 1 [40,43,46] outlines the process of intervention development. PEDAL2 interviews were conducted between August 2019 and November 2020, and the findings from these interviews and the development of the conceptual framework are reported elsewhere [43]. Intervention development, including PPI discussions with individuals living with breast or prostate cancer, expert review and feedback, and instructor workshops and feedback, took place between September 2021 and March 2022. Patient partners with breast cancer (n=4) were recruited through an existing patient and partner group established for a wider research program, while patient partners with prostate cancer (n=6) were recruited through a local prostate cancer charity. Instructors (n=5) were recruited through Life Cycle, the community organization involved in delivering an e-cycling intervention to another clinical population (PEDAL2) [40].

The intervention was refined following feedback from CRANK PPI members (n=10), instructors (n=5), and the expert group (n=3). Specifically, the intervention material was sent to PPI members initially for review. After review of these documents, group members met on the web, with 3 group discussions for each of the 3 groups, where feedback on the intervention content and delivery method was provided to the research team. All discussions were recorded in order for the researcher to listen back to the discussion, and notes were taken throughout the discussions. These discussions lasted up to 120 minutes. For instructors, the intervention material was sent for review prior to a 2-hour web-based meeting. At this meeting, the research team presented the intervention material, and instructors were asked to provide feedback on both the instructor training and participant intervention material. Any disagreements within the groups were discussed in the session until a consensus was reached regarding the required intervention changes. The intervention was sent to members of the CRANK trial steering committee with expertise in behavior change. These individuals provided written feedback on the intervention content. The information from CRANK PPI members, instructors, and the trial steering committee was collated and reviewed by 2 researchers, and consensus decisions were made on appropriate changes to be made based on the APEASE criteria.

The National Health Service Health Research Authority Southwest/Central Bristol Research Ethics Committee provided ethics approval to conduct qualitative interviews among individuals living with T2D as part of PEDAL2 (reference 18/SW/0164). While ethical practice was upheld, formal ethical approval was not obtained for these PPI activities, as advised by the National Institute for Health Research. PPI contributors were involved in the design, implementation, and management of the research process itself. Patient partners were informed of what engagement in the PPI activities would entail prior to agreeing to participate. Ethics approval for the pilot randomized controlled trials to evaluate the intervention (named CRANK) was granted by the National Health Service Health Research Authority Dulwich Research Ethics Committee (REC: 22/LO/0036; CRANK-P) and the Nottingham Research Ethics Committee (REC: 22/EM/0010; CRANK-B), and the protocol for this study is reported by Bourne et al [50].

Justification for this research is provided in the Introduction section and summarized here for completeness. Individuals living with cancer are less physically active than their healthy counterparts [15,17,51,52]. Specifically, it has been estimated that adherence to the physical activity guidelines among individuals living with cancer ranges from 10% to 47% [15,16,52-56]. Engagement in regular physical activity can help recovery from cancer, reduce the side effects associated with treatment, and reduce the chances of recurrence [3,5,10,11,57]. Several physical activity interventions have been developed for individuals living with and recovering from cancer, with varied success [58-61]. Common barriers to engagement in regular physical activity for individuals with cancer include cancer-related physical symptoms (eg, fatigue), lack of equipment or facilities, lack of knowledge of what to do or support or advice on how to engage in physical activity, lack of motivation and time, and low confidence [21-25].

e-Cycling has been identified as an activity that may overcome some of the identified barriers to engaging in physical activity and promote long-term adherence [28]. The potential of e-cycling to increase physical activity in individuals being treated for or recovering from cancer has not been explored but warrants further investigation. As such, the target behavior of this intervention was to increase physical activity specifically through e-cycling in individuals being treated for breast cancer or prostate cancer.

The aim was to increase individual’s weekly physical activity through engagement in e-cycling. No specific weekly e-cycling targets were imposed by the research team, as we wanted participants to have autonomy over their e-cycling goals. The e-cycling behavior will need to be performed by individuals aged 18 years or older with a diagnosis of prostate cancer or breast cancer on a regular basis, whenever possible with the mantra that every move counts. e-Cycling can be conducted for any purpose (ie, exercise, travel, social, and leisure) in an outdoor setting. e-Cycling can be performed alone or with others.

Drawing on the conceptual model developed from PEDAL2 and incorporating information from PPI discussions with patients being treated for breast cancer or prostate cancer, the intervention team identified key mechanisms of action to target in the intervention to bring about engagement in e-cycling (Tables 1-3).

In addition, a review of the literature identified the use of Social Cognitive Theory (SCT) and Transtheoretical Model of Behavior Change as appropriate theoretical models to explain why people adopt physical activity behavior, particularly those with breast cancer and prostate cancer [60-63]. The key constructs of these theoretical models were considered when identifying what needs to change for e-cycling to take place. SCT uses the techniques of mastery, vicarious experiences, and modeling to develop skills and build self-efficacy [64]. SCT also highlights the importance of others when changing behaviors. The Transtheoretical Model is a comprehensive model of behavior change [65]. The 10 processes of change focus on “how” individuals change their behavior. In addition, 2 intervening variables of self-efficacy and decisional balance have been identified as impacting movement between the 6 stages of change.

Overall, 11 of the 14 domains of the TDF were identified as needing to be targeted to encourage engagement in e-cycling. These are shown in Tables 1-3 and summarized below. Specifically, ensuring individuals had the physical skill and knowledge to ride the e-bike and navigate traffic was identified as essential for e-cycling engagement, as was having high confidence to ride the e-bike itself and among traffic (belief about capabilities). To compete with the car as a mode of transport, individuals noted that systems must be in place to ensure e-cycling is perceived as the “easy” option (memory, attention, and decision processes) and that establishing a routine was key (reinforcement). In addition, ensuring individuals have the correct equipment and access to a breakdown service would facilitate e-cycling engagement (environmental context and resources). Setting goals, monitoring the process toward these goals (behavioral regulation), and encouraging individuals to seek out support from family and friends (social support) were seen as important factors that will increase the likelihood of e-cycling behavior. Furthermore, holding positive beliefs about the impact of e-cycling both in regard to physical and mental health and social and environmental outcomes (belief about consequences; emotions) was important to influence behavioral engagement.

A total of 23 BCTs linked to the theoretical domains, as identified through the theory and techniques of web-based tool [46], psychological theories, and literature on BCTs effective at increasing physical activity among individuals being treated for breast cancer or prostate cancer were identified as shown in Tables 1-3. There was significant overlap in the BCTs identified as potentially useful to target the underlying mechanisms of change. The techniques fall across 12 of the 16 BCT categories of goals and planning, feedback and monitoring, social support, shaping knowledge, natural consequences, comparison of behavior, associations, repetition and substitutions, rewards, regulation, antecedents, and self-belief.

The intervention was refined based on feedback from PPI members, cycling instructors, and experts.

This study reports on the development of a behavioral intervention designed to increase e-cycling behavior in individuals living with prostate cancer or breast cancer. The systematic approach to intervention development involved drawing on theory, evidence, and end-user insights to identify appropriate factors to target in this intervention. It is believed that this process will increase the potential efficacy of the intervention and will allow for an in-depth evaluation of the intervention content to gain an understanding of the intervention effects.

In total, 11 of the 14 domains of the TDF, covering all components of the COM-B model, were found to be relevant to increasing physical activity behavior through e-cycling among individuals living with prostate cancer or breast cancer. Constructs of the TDF not targeted in this intervention were professional or role identity, optimism, and intentions. The 11 TDF constructs identified will be targeted through the use of 23 BCTs from 12 overarching BCT categories [47]. The categories of BCTs selected for use in this intervention are similar to those identified by McVicar et al [67] in their development of an e-cycling intervention for overweight and obese adults using participant workshops. In this intervention, the categories of rewards and regulations were incorporated, which were not part of the intervention developed by McVicar et al [67]. Specifically, in this intervention, individuals were encouraged to reward themselves, primarily through self-praise if they met, or made progress toward, their e-cycling goals. The use of self-rewards has been associated with sustained physical activity behavior at least 6 months after intervention [68]. In addition, participants were encouraged to prepare cycling equipment ahead of time to reduce stress and increase the likelihood of e-cycling over using a motorized vehicle. Ways to assist with behavioral regulation were included in this intervention, as interviews with PEDAL2 participants revealed that trying to remember everything needed for a commute via e-bike was stressful and decreased the likelihood of riding. Overall, McVicar et al [67] identified 16 BCTs for inclusion in their intervention, while the current development process identified 23, of which 12 overlapped. The additional 11 unique BCTs used in this intervention were likely due to the incorporation of sources of information that identified the mechanisms of action that broadly impact PA behavior in the current clinical population as well as those that impact e-cycling specifically. For example, BCT 15.3 focus on past success was incorporated as this technique directly aligns with bolstering self-efficacy, a key component of SCT [64], which has been used in previous physical activity interventions among individuals living with cancer. In addition, BCT 6.5 information about emotional consequences was included due to the conceptual model of e-cycling engagement among individuals with T2D and the finding that individuals are more likely to engage in e-cycling because it is perceived as enjoyable [43]. The BCT 2.3 self-monitoring was found to be an important component to prompt behavior change from a theoretical perspective, from the PEDAL2 conceptual model, and based on discussions with end users and was therefore included in this intervention.

Several of the techniques identified for use in this intervention align with action types identified by Kelly et al [69] in a scoping review of 145 initiatives reporting on intervention content, at the organization and individual levels, aimed at increasing cycling behavior. The review identified commonly used action categories, which will also be used in this intervention. These include knowledge of the benefits of cycling and cycle safety and route planning (ie, education), practical cycling training courses (ie, training), provision of bike accessories and bikes, and access to bike maintenance services (ie, enablement). While the review identified a series of actions associated with restructuring the environment, the majority of these were not suitable for this intervention (eg, provide bike storage facilities and bike wheel channels on stairs or workplace or organizational policies). These components are likely more achievable for organization-based interventions.

The behavioral analysis conducted in this study outlines the systematic process used to develop a theoretical understanding of the behavior we are seeking to impact and the mechanisms that may influence this behavior. From here, we were able to identify the theoretical constructs to target and the techniques through which to target these mechanisms in the current population. This transparent method demonstrates the multifactorial nature of this behavior and the complexity of developing a behavioral intervention. However, documenting this process is important, as it allows others to fully understand how the active ingredients of the intervention were selected.

To gain an in-depth understanding of the behavior of e-cycling and the factors that influence engagement, we drew on theory and literature and engaged stakeholders including individuals living with prostate cancer or breast cancer, cycling instructors, and experts in the field of behavior change. It is hoped that gathering information from multiple sources to guide intervention development will increase the chances of developing an intervention that can effectively increase physical activity behavior through e-cycling participation.

The BCTs selected to target each TDF construct were selected using the links proposed by the Theory and Techniques tool, which links BCTs and mechanisms of action based on evidence in the literature [70] and expert consensus [71] and triangulation of these 2 processes [46]. The use of this tool is more appropriate than the use of the BCW guidance, which links BCTs to mechanisms of action based on the “most used” techniques [32].

A potential limitation of this intervention development is that the conceptual model used to guide this intervention was based on findings from 1 city, the same city in which this intervention will be delivered (Bristol, United Kingdom). While appropriate for this intervention, mechanisms of change identified and associated intervention active components may not be applicable to individuals from other cities in the United Kingdom or internationally. A second limitation is that one intervention has been designed for 2 clinical populations, specifically breast cancer and prostate cancer. These cancers generally impact different genders, and there is the potential that these individuals have different barriers and facilitators to e-cycling engagement that may not have been parsed out in this process. However, this work included PPI discussions with both individuals living with breast cancer or prostate cancer, and no outstanding differences were noted between the 2 PPI groups.

This process has led to the development of an intervention with associated participant intervention materials to address some of the barriers identified to e-cycling engagement. In addition, an instructor manual has been created to ensure instructors address these barriers and engage in activities that facilitate e-cycling through training and discussion. The intervention package is currently being tested in a pilot randomized controlled trial [50]. Specifically, the feasibility of delivering this intervention and specific BCTs is being assessed through observations of sessions delivered by instructors with feedback as well as intervention checklists completed during each contact session. The frequency with which each BCT is delivered will be determined and reported. Through workbooks and self-monitoring tools, we will be able to establish the extent to which participants engaged with the BCTs. The impact of omission of BCTs can be compared to effectiveness data and may give insight into the potential efficacy of individual or groups of BCTs selected. In addition, qualitative one-to-one interviews will be conducted with instructors and participants to understand the acceptability of intervention delivery and participation. Testing the delivery of the intervention components is a critical part of intervention development to ensure refinements can be made where required prior to full-scale implementation if suitable. The intervention developed, if appropriate, can be adapted to other groups of individuals being treated for different cancers using the most potent BCTs identified.

This study presents the process of designing a behavior change intervention targeting physical activity behavior using electrically assisted bicycles for individuals living with breast cancer or prostate cancer. The explicit reporting of the development process and program theory with associated intervention content facilitates the application of in-depth evaluation to determine the efficacy of the BCTs included. This evaluation is currently being conducted and will enable future refinement of the intervention as appropriate.