Activities that have been carried out as part of the development of the Renewed intervention are reported in more detail in the following publications:
Corbett T, Singh K, Payne L, Bradbury K, Foster C, Watson E, et al. Understanding acceptability of and engagement with Web-based interventions aiming to improve quality of life in cancer survivors: a synthesis of current research. Psycho-Oncology 2018;27:22–33. https://eprints.soton.ac.uk/414881/
Bradbury K, Steele M, Corbett T, Geraghty AWA, Krusche A, Heber E, et al. Developing a digital intervention for cancer survivors: an evidence-, theory- and person-based approach. NPJ Digit Med 2019;2:85. https://doi.org/10.1038/s41746-019-0163-4
Corbett T, Cheetham T, Muller AM, Slodkowska-Barabasz J, Wilde L, Krusche A, et al. Exploring cancer survivors’ views of health behaviour change: ‘where do you start, where do you stop with everything?’. Psychooncology 2018;27:1816–24. https://eprints.soton.ac.uk/419278/
Introduction
The development process for creating the Renewed intervention was based on a combined evidence-, theory- and person-based approach.11–13 The development work focused primarily on developing introductory content and two components to support physical activity (‘Getting Active’) and healthy eating (‘Eat for Health’). In addition, two of our existing digital interventions (POWeR + for weight management14 and Healthy Paths for distress management15) were adapted and included in the intervention.
Intervention development team and patient and public involvement
The intervention development team included patient and public involvement (PPI) representatives, clinicians with expertise in supporting those who have experienced cancer, general practitioners (GPs), behavioural scientists, experts in digital health, representatives of the charities Breast Cancer Now, Bowel Cancer UK and Prostate Cancer UK. Our PPI contributors comprised two patients who had experienced breast cancer (Tamsin Burford, Lesley Turner), two patients who had experienced bowel cancer (Peter Roberts, Kevin Summers) and two patients who had experienced prostate cancer (Geoffrey Sharman, Roger Bacon). Some of the PPI members were also involved in work with additional cancer charities (e.g. Macmillan, Prostate Cancer Support Organisation). All PPI members joined the project team and were invited to join monthly development group meetings to discuss the intervention development. They were also invited to quarterly management committee meetings to discuss important issues within the planning and development of the overall programme of research. Within the monthly development group meetings, PPI contributed to all discussions and decisions regarding the key intervention content. For example, PPI informed decisions about the frequency of human support that would be offered alongside the digital intervention and the key points at which this should be available. Within the quarterly management, meetings PPI contributed to decisions about the running of the research (such as how to optimise recruitment). PPI also commented on all patient facing recruitment materials to optimise the accessibility and persuasiveness of these materials.
The PPI group and clinicians commented on early versions of the written website content to ensure this was as accessible, feasible, motivating and persuasive as possible. PPI also each took part in three think-aloud interviews which helped to optimise this content further. For example, they suggested additional concerns about increasing physical activity that patients might have, which the research team were then able to address within the intervention. They also highlighted some areas of the digital intervention where navigation could be simplified to maximise accessibility.
Overall, PPI contributors promoted a focus on patent priorities throughout the intervention planning and optimisation. Their rapid feedback on early prototype intervention materials enabled the research team to make these as useful as possible before they were viewed by patients recruited from primary care in the qualitative think-aloud interviews.
Workstream 1 and workstream 2 objectives
This section describes the three phases within which activities were carried out to inform the planning and development of the Renewed intervention. Each activity is described, reporting on aims, methods used, results, and their practical implications for intervention development. The phases were:
Workstream 1 (phase 1): literature reviews
- Thematic synthesis of qualitative evidence
- Rapid scoping review
Workstream 1 (phase 2): intervention planning
- Guiding Principles
- Behavioural analysis
- Logic model
Workstream 2 (phase 3): intervention development and optimisation
- Qualitative research with cancer survivors
- Qualitative research with NHS and cancer charity staff and volunteers
This section ends with a description of how the recommended IdentifyiNg and assessing different approaches to DEveloping compleX interventions (INDEX) actions for developing complex interventions16 were considered and addressed during the planning and development phases of the Renewed intervention.
Workstream 1 (phase 1): collating evidence from reviews of the literature
Thematic synthesis of the qualitative evidence for digital interventions for lifestyle and well-being support for cancer survivors (described in Corbett et al.17)
Aims
To identify features and outcomes of web-based interventions which may be important to the quality of life of cancer survivors.
Methods
A rapid review was conducted to identify studies related to web-based interventions designed to improve quality of life in adults who have completed primary treatment for breast, prostate and colorectal cancer. A range of study designs were included (qualitative research, feasibility/pilot trials, randomised trials and process evaluations). All available information about intervention characteristics, experiences and outcomes was extracted. The data were treated as textual (qualitative) data and analysed using thematic analysis.
Results
The database search identified 6327 papers, of which 57 were assessed for eligibility through full-text screening. A total of 16 relevant papers relating to 9 interventions fulfilled all eligibility criteria for inclusion. Twenty-eight descriptive themes were identified and grouped into five analytical themes.1 Four themes addressed aspects of intervention designs and implementation of web-based intervention.18 The themes were: participant factors, characteristics of the online intervention, techniques used to change behaviour and preferred features of web-based interventions. These themes were perceived as key factors that appeared to potentially influence a fifth analytical theme: the outcomes discussed in the papers including uptake, adherence and attrition, engagement, feasibility, efficacy, positive behaviour change and acceptability of the interventions. The results of this review have been published.17
Practical implications for intervention development
The findings highlight the importance of matching the intervention to the unique characteristics of the participants. The key findings suggest that survivors value interventions that consider their changing needs and are tailored to their stage of the cancer trajectory. Social networking features do not always provide added benefit and it is likely that web-based interventions will be effective without social networking features and with little input from clinical staff and researchers. Behavioural change techniques used in interventions need to be appropriate to the characteristics of the users and implemented carefully to avoid negative consequences. This information is vital in the development of successful interventions for cancer survivors.
Rapid scoping review to identify barriers and facilitators (described in Bradbury et al.19)
Aims
To review existing evidence to identify facilitators and barriers to the success of interventions designed to improve the quality of life of cancer survivors, in order to inform the development of Guiding Principles,11 theory-based behavioural analysis and logic modelling20 for the Renewed intervention.
Methods
Five of the six core steps described by Arksey and O’Malley for carrying out scoping reviews were followed: identifying the research question, identifying relevant studies, study selection, charting the data, and collating, summarising and reporting the results.21 The optional sixth step (a consultation exercise to seek expert consensus), was not carried out due to time constraints, although feedback from PPI representatives and members of the expert development group was sought.
A search was conducted using the Cochrane Library, Database of Abstracts of Reviews of Effects, Ovid MEDLINE and PsycInfo® (American Psychological Association, Washington, DC, USA), limiting the search to the past 20 years. Details of the inclusion criteria used for selection of studies have been published,19 along with a Preferred Reporting Items for Systematic Reviews and Meta-Analyses22 flow diagram, summarising the rapid scoping review, which also included further studies identified by experts in the team and through examination of reference lists of identified studies. This process yielded 49 studies to be retained for analysis, which resulted in a list of participant and intervention component characteristics for which potential facilitators and barriers were identified.
Results
Facilitators and barriers were grouped according to nine participant and intervention characteristics, namely, factors influencing participation (e.g. to protect against recurrence of cancer), information included in the intervention (e.g. including topics that are relevant to users), motivation/self-esteem/self-efficacy (e.g. lack of confidence or motivation to engage in suggested activities), self-management and self-monitoring methods (e.g. having a sense of having control over one’s body), emotions/mood (e.g. fear of recurrence of cancer), social support (e.g. availability or lack of support, and its effect morale), intervention design/content (e.g. relatability, tailoring), technical aspects (e.g. navigation difficulties, ability to print information) and various practical issues (e.g. safety issues, time constraints, bad weather). The full list of facilitators and barriers was published.19
Practical implications for intervention development
This rapid scoping review highlighted several facilitators and barriers to take into consideration for the success of interventions aiming to improve quality of life in cancer survivors, which were then used to inform the guiding principles, behavioural analysis, and logic model, which in turn informed the design and content of the intervention. These are described in phase 2.
Workstream 1 (phase 2): intervention planning
Guiding Principles: identifying design objectives to meet target users’ needs and maximise engagement (described in Bradbury et al.19)
Aims
To develop Guiding Principles11 which outline key behavioural issues and needs of target users, and how these will be addressed by the Renewed intervention in terms of design objectives and key intervention features.
Methods
Findings from the rapid scoping review (described in phase 1) were used to identify key context-specific behavioural issues, in terms of target user needs, issues, or challenges. For each of these, intervention design objectives were formulated, describing what the intervention will aim to do to address behavioural issues in order to maximise engagement with the intervention and the behavioural changes suggested. Finally, key intervention features were developed to describe how the intervention will achieve each design objective. These Guiding Principles were improved over time in consultation with members of the development team and of our stakeholder panel who have extensive expertise of cancer survivorship and knowledge of relevant literature.
Results
One key issue related to users not seeing themselves as having health needs or requiring health lifestyle changes after cancer, which could compromise engagement with the healthy lifestyle changes put forward by the intervention. To prevent this scenario, the preferred approach to presenting content would be to promote well-being, rather than suggesting ways to manage illness. Another issue related to a risk that cancer survivors could feel their usual behaviours, when compared with the intervention’s proposed behaviours, were inappropriate, and that they may have even had a part to play in causing their cancer. The design objective for this issue consisted of ensuring that any suggestions for behavioural change did not stigmatise users’ current behaviour, which could be achieved by avoiding using arguments or ways of presenting certain activities that could result in users feeling blamed for their cancer or for their poor mental health (e.g. overpromoting ‘positive coping’). A third issue was that target users form a heterogeneous group in terms of the wide variety of symptoms and issues they experience, depending on cancer type, gender and individual circumstances. To address this, one of the key intervention features was to provide tailored information to each user, based on answers given as part of baseline quality of life measures, as well as gender and cancer type, so that it is as relevant as possible to them. Another issue related to the importance of information is being brief and easily accessible. Examples of intervention features for achieving this were to provide short sessions on specific topics that were also accessible via smartphones, with implications for the amount of text to be presented. The last Guiding Principle refers to the risk of the intervention being too large and complex as a result of trying to address multiple symptoms and proposing many behaviours to address them. To avoid this, an efficient design was used, for example by targeting behaviours which had the potential to improve multiple symptoms. The Guiding Principles were published.19
Practical implications for intervention development
The Guiding Principles provided a useful resource for development team members to refer to as a reminder of what challenges the intervention’s target population face when engaging with its content, and how the intervention design, content, features, presentation approach, and tone, should address these challenges.
Behavioural analysis: identifying barriers, and how to address them (described in Bradbury et al.19)
Aims
To describe the Renewed intervention and potential determinants of behaviour for each target behaviour, integrating theoretical frameworks from behavioural change and implementation sciences.
Methods
A behavioural analysis table was populated with likely barriers to achieving each target behaviour. These barriers were extracted from the scoping review findings described in phase 1, as well as from expert consultation. For each barrier, intervention components designed to minimise them were described and coded according to three theoretical frameworks, namely, the Behaviour Change Techniques Taxonomy (BCTv1),23 the Behaviour Change Wheel (BCW)24 and Normalisation Process Theory (NPT).25
Results
Three target behaviours were identified: increase in physical activity, adopting a healthier diet and patient engagement with the intervention. For each of these behaviours, a range of barriers were listed, along with specific intervention components that would minimise them, as well as codes describing how these were mapped to elements of the BCTv1, BCW and NPT theoretical frameworks. For example, target users’ concerns about being physically active were addressed by presenting success stories of other individuals like them, describing how they overcame concerns about being more active. This intervention component targeted three BCW constructs (psychological capability, reflective motivation and social opportunity), two BCW intervention functions (persuasion and modelling), one NPT construct (coherence) and three BCTv1 behaviour change techniques (9.1 credible source, 3.1 social support and 5.6 information about emotional consequences). The complete behavioural analysis table has been published.19
Practical implications for intervention development
The behavioural analysis was useful to ensure the intervention addressed as many known barriers as possible to achieve the intervention’s key target behaviours. Although the mapping of barriers and intervention components to theoretical frameworks did not identify additional ones, it helped with ensuring none had been overlooked.
Logic model: outlining the proposed mechanism of action of the intervention (described in Bradbury et al.19)
Aims
To outline a model of the hypothesised mechanisms of action of the Renewed intervention (how it is thought to work), which can then be later tested and refined during the process analysis.
Methods
Drawing on Medical Research Council guidance for process evaluation available at the time,26 the behavioural analysis and findings from the rapid scoping review (described in phase 1) were used to develop the intervention’s logic model,20 which describes the problem the intervention addresses, intervention targets to resolve the problem, intervention ingredients, mechanisms of action, and outcomes the intervention is aiming to impact.
Results
All elements of the logic model were described, with poor quality of life in cancer survivors being the problem to be addressed by the intervention. Adopting healthy behaviours and improving mental health were identified as the intervention’s targets. Intervention ingredients were defined, including, for each, details of the psychological constructs they target, as outlined in the behavioural analysis table. This is followed by a description of the proposed mechanisms of action and outcomes in a flow diagram format. For example, it is hypothesised that engagement with the intervention would result in the desired behaviour changes (increasing physical activity, improving dietary patterns), which would in turn improve quality of life (the primary outcome) and reduce anxiety and depression, fear of cancer recurrence and improvements in well-being and enablement (the secondary outcomes). The full logic model was published.19
Practical implications for intervention development
The logic model was useful for defining how the intervention is expected to lead to its effects and for providing a further check that the intervention design features and component are appropriate in terms of supporting the proposed mechanisms of action.
Workstream 2 (phase 3): intervention development and optimisation
Qualitative research: think-aloud interviews with cancer survivors (described in Bradbury et al.19)
Aims
To explore target users’ views on the prototype of the Renewed intervention, in order to identify modifications to be made to the intervention with a view to optimising user engagement with both the intervention and the behaviour changes it suggests.
Methods
Thirty-two cancer survivors who had completed treatment for breast, colorectal or prostate cancer in the past 10 years took part in the study by completing three separate think-aloud interviews each on different parts of the intervention. These interviews consisted of participants using the intervention for the first time and voicing their immediate reactions out loud in the presence of a trained interviewer who audio recorded the interview. This was followed by brief semi-structured interview questions to gain further insight into what they thought of the intervention as a whole, including what they thought about the behaviour changes it suggests. Positive and negative comments were then collated in a table, grouped by page, section or intervention feature. The table was then reviewed by the wider development team (including PPI representatives, clinical cancer experts and charity members), with potential modifications to the intervention discussed and agreed, especially if they related to comments that had been voiced by more than one participant, if they were critical to behaviour change, and were in line with the intervention’s Guiding Principles. Further interviews were then conducted with a new modified version of the prototype. This iterative process continued until no important further changes were required.
Results
In terms of positive comments, participants found the intervention to be generally easy to navigate, also commenting on liking the look of the intervention. They mentioned valuing the fact that it was designed by experts in the field, and that this increased how much trust they could put in the information it provides. Most importantly, participants also commented on the content being relevant and useful to them.
Several negative comments describing barriers to engagement were also made, which resulted in modifications to be made to the intervention prototype. For example, some participants explained that they were worried about overdoing physical activity if they were to increase it. To address this concern, some changes were implemented. These included emphasising that this is a common concern, as well as suggesting that increasing physical activity should be done slowly and gradually, in order to avoid the more negative consequences of overdoing activity. The entire physical activity section was reviewed to ensure that all messages about increasing physical activity mentioned this should be done slowly.
Another example of an important barrier to behaviour change relates to the advice about reducing the consumption of meat in the healthy eating section of the intervention, which was perceived as promoting an inappropriately radical change in diet. This was not well received by all, with mentions of not liking the idea of having meat-free meals. This resulted in removing this suggestion and to instead encourage users to reduce the amount of red and processed meats they consume. Related to this, some participants found the argument about eating less meat benefiting the planet very off-putting. As a result of this finding, it was removed from the intervention. An overview of all the barriers and changes that have been made to the intervention to address them has been published, along with some examples of participant feedback.19
A thematic analysis of discussions on broader issues that took place during these interviews was also carried out. An important conclusion, which was published,18 relates to whether this period of time after cancer treatment constitutes a ‘teachable moment’, a time during which individuals become motivated to make healthy lifestyle changes.27 Most participants mentioned several general and cancer-specific barriers to making lifestyle changes after successful cancer treatment, which suggests that for the sample of participants we interviewed, this period of time did not appear to clearly constitute a teachable moment. This important question will be revisited as part of the qualitative process study to determine the reasons why the intervention may have not been effective for everyone.
Practical implications for intervention development
This qualitative research proved invaluable for understanding which aspects of the intervention target users liked and which potentially required modifications. This allowed the development team to optimise the intervention in order to maximise engagement and the behaviour changes it suggests.
Qualitative research: focus groups with National Health Service and cancer charity staff and volunteers (described in Bradbury et al.19)
Aims
To explore potential supporters’ views on the proposed support to be provided to users of the Renewed intervention, as well as the online training provided. To use findings to optimise the training material, the delivery of support, and to identify the most suitable supporters.
Methods
Seven focus groups were carried out with staff from five GP practices, as well as with staff and volunteers from two cancer charities. Topics for discussion included current support provisions, views on using an online tool for support, the online training material, the proposed CARE (Congratulate, Ask, Reassure, Encourage) approach (detailed elsewhere28), fitting in the intervention and supporter role in everyday practice and the use of e-mail to communicate with patients. The schedule that was used for these focus groups has been published.19 Optimisation of the supporter aspect of the study was an iterative process whereby feedback was recorded in a table and discussed within the development team, after which agreed modifications to the training and support procedures were implemented. Additional focus groups were carried out until no further improvements were identified.
Results
Several concerns were raised which required modifications. For example, participants expressed unease with some of wording that was used in the online training, which discussed how the intervention aimed to improve quality of life and prevent cancer recurrence. Some NHS staff worried that as a result some supporters might give assurances to users of the intervention that they would not get cancer again if they followed all the advice. The development team agreed that this would pose a risk, and as a result rephrased ‘prevent cancer recurrence’ to ‘lower chances of recurrence’ in the supporter training material.
There were also some questions around the use of the CARE approach from some NHS staff. For example, how to apply the ‘congratulate’ part when providing support to a patient who is not achieving their goals. As a result, additional information about how to provide support with patients who are not successful was added to the guidelines on how to use the CARE approach.
Another concern related to ensuring that intervention users would not be left to feel abandoned should a supporter not respond to them due to being away because of unexpected long-term illness or holidays. As a result, the intervention was modified to enable GP practices to nominate a second supporter who could, if required, complete the training and take over as a supporter.
Discussions around which individuals would be suitable to take up the supporter role also took place. GP practice staff felt that healthcare assistants (HCAs) would be able to provide the support as no medical knowledge is required for this role. As a result, a decision was made to aim to recruit HCAs for the trial, and to also allow nurses to provide support as well as required.
An overview of the all the concerns and changes that have been made to the intervention to address them has been published, along with some examples of participant feedback.19
Practical implications for intervention development
Findings from these focus groups enabled the development team to refine the online training material (and therefore improving the quality of the support), practical aspects of the delivery of the support, as well as identifying suitable staff from GP practices who could conduct the role of supporter.
Mapping the Renewed planning and development process to the IdentifyiNg and assessing different approaches to DEveloping compleX interventions actions16
The INDEX actions provide guidance on the development of complex interventions to improve health and health care. The authors developed the INDEX actions based on a consensus exercise which was informed by reviews and qualitative interviews with developers and wider stakeholders. Appendix 1 shows the retrospective mapping of the Renewed planning and development process to the 11 INDEX guidance actions.
A full description of the Renewed intervention using the Template for Intervention Description and Replication (TIDieR) checklist29 is available in Report Supplementary Material 1.
Workstream 1 and workstream 2 limitations
Due to time constraints, the rapid scoping review was limited to literature published in the past 20 years in three databases. As such, the grey literature was not searched, potentially resulting in some literature findings being overlooked. The optional sixth step described by Arksey and O’Malley for carrying scoping reviews,21 a consultation exercise to seek expert consensus, was not included. As we had consulted our expert and PPI development group, we did not feel that this step was sufficiently high priority to carry out within our limited time frame. This might have potentially resulted in more refined findings.
Men and ethnic minorities were underrepresented in the focus groups that took place with NHS and cancer charity staff and volunteers. As such, it is possible that their views were different to those included in our findings.
Finally, for both qualitative optimisation studies, it is possible that those who chose to not take part in them had different views to those expressed by those who did.
Workstream 1 and workstream 2 conclusions
The activities that have been carried out as part of the intervention development phase helped ensure that the Renewed intervention was accessible, acceptable, persuasive, motivating, and feasible to implement in practice for cancer survivors.