The use of cardiac rehabilitation services to aid the recovery of patients with bowel cancer: a pilot randomised controlled trial with embedded feasibility study

Article history

The research reported in this issue of the journal was funded by the HS&DR programme or one of its preceding programmes as project number 12/5001/09. The contractual start date was in June 2013. The final report began editorial review in June 2015 and was accepted for publication in January 2016. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HS&DR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

Shaun Treweek reports a grant from the National Institute for Health Research during the conduct of the study.

Copyright statement

© Queen’s Printer and Controller of HMSO 2016. This work was produced by Hubbard et al. under the terms of a commissioning contract issued by the Secretary of State for Health. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

Rationale for cancer rehabilitation

Increasing 5- and 10-year survival rates mean that many cancers are now considered chronic diseases. There are approximately 28 million people living with and beyond cancer in the world1 and many of these cancer survivors face ongoing challenges from the post-treatment care standpoint. In this report, we use that term, cancer survivor, to refer to someone who is living with and beyond cancer; we recognise, however, that not all cancer survivors would identify themselves using the term ‘survivor’. 2 In the UK, there are over 2 million people living with and beyond cancer, a figure which is rising by 3% per annum. 3 As such, supporting cancer survivors represents one of the largest UK and global health challenges.

Colorectal cancer is also called bowel cancer and includes large bowel cancer (colon cancer) and cancer of the back passage (rectal cancer or cancer of the rectum). CRC is the second most frequently diagnosed cancer in women and the third most frequently diagnosed cancer in men, accounting for 1.23 million new CRC cases in 2008 worldwide. 4 CRC is the fourth most common cancer in the UK and there are approximately 244,000 people living with and beyond CRC. 5 In 2011 in the UK, 41,581 people were diagnosed with CRC (13% of all cancer cases); 57% of adults who were diagnosed with CRC (56% of men and 57% of women) in 2010–11 in England and Wales were predicted to survive ≥ 10 years. 6 Addressing the post-treatment needs of this group is, therefore, a UK public health priority.

Colorectal cancer survivors report ongoing and persistent physical and psychological impairments. 7–9 Physical symptoms include fatigue, physical discomfort and bowel function problems (e.g. diarrhoea, frequency of bowel movement and incontinence), and these may be present in up to 72% of survivors. 9 The clinical rationale for cancer rehabilitation is to support the management of late and long-term effects of cancer and its treatment, increase chances of survival and improve general health and quality of life.

The American Cancer Society and the World Cancer Research Fund recommend that cancer survivors would benefit from following lifestyle recommendations for secondary cancer risk reduction (e.g. taking a nutrient-dense diet, increasing levels of physical activity, ceasing smoking, reducing alcohol intake and avoiding excess body fat). 10,11 The adoption of lifestyle recommendations may also reduce CRC survivors’ risk of other diseases, such as cardiovascular disease, which is a common comorbidity in people diagnosed with CRC,12 the aetiology of which is also attributed to lifestyle factors. 13

However, most CRC survivors are not meeting these lifestyle recommendations. 14–18 A study conducted in Australia, for instance, shows that, at 12 months post diagnosis, approximately 8% of CRC survivors are smokers, 22% are high-risk drinkers, 62% are insufficiently physically active and 61% are overweight/obese. 14 CRC survivors have been found to have the lowest physical activity rates of any cancer group. 19 There is a strong case, therefore, for the provision of rehabilitation for cancer survivors, and, in particular, for people with CRC. However, at the time the protocol for this current study was written (2012), cancer rehabilitation was not usual care in the UK or, indeed, elsewhere. 20 A challenge facing the NHS, therefore, is integrating rehabilitation into standardised models of care for cancer survivors.

The National Institute for Health and Care Excellence guidance, Improving Supportive and Palliative Care for Adults with Cancer, states that:

Rehabilitation attempts to maximise patients’ ability to function, to promote their independence and to help them adapt to their condition. It offers a major route to improving their quality of life, no matter how long or short the timescale. It aims to maximise dignity and reduce the extent to which cancer interferes with an individual’s physical, psychosocial and economic functioning. 21

Cancer rehabilitation is a care specialty that comprises the full spectrum of rehabilitation fields, including the physical, psychosocial and socioeconomic, and can include physical activity, diet, nutrition and psychosocial components. 22 As we have already pointed out, cancer rehabilitation is an often-neglected aspect of cancer care in terms of health policy and infrastructure. 20 Reasons for this include clinicians remaining unconvinced or unaware of evidence of patient benefit from rehabilitation, and challenges around the implementation of rehabilitation in current cancer care pathways. The following section explains why rehabilitation for people with CRC should include physical activity as a core component.

Evidence for increasing physical activity

Physical activity is a key component of rehabilitation. In order for clinicians to prescribe physical activity, which is a non-pharmacological adjunctive therapy, for CRC survivors, there needs to be strong evidence of patient benefit. The clinical rationale for physical activity interventions for CRC survivors is derived from epidemiological observations of relationships between physical activity and survival, and evidence of cause and effect derived from randomised controlled trials (RCTs) about the benefits of physical activity on psychosocial domains, such as quality of life, fatigue, anxiety and depression. 23,24 We use the term ‘physical activity’ throughout this report, although we recognise that the term ‘exercise’ is often used interchangeably to refer to ‘A potential disruption to homeostasis by muscle activity that is either exclusively or in combination, concentric, eccentric or isometric’. 25

To assist the reader in interpreting the following evidence about associations between physical activity and clinical endpoints in people with CRC, and, perhaps more importantly, to recognise the level of physical activity required to achieve a health benefit, we describe current recommended UK guidance for physical activity for the general population and metabolic equivalents (METs) in Boxes 126 and 2, respectively. 27

Physical activity and survival

Three separate meta-analyses23,28,29 including the same six observational studies30–35 were published in 2013–14. Owing to different cut-off values for level of physical activity and different statistical analyses used, there is slight variation in the results; what is evident, however, is that all three studies show that a higher level of physical activity is associated with an increase in cancer-specific and overall survival.

A meta-analysis28 reported that three32,34,35 out of the six prospective cohorts assessing post-diagnosis physical activity found a statistically significant increase in cancer-specific survival among patients with a high level of physical activity, compared with patients with a low level. Overall, higher post-diagnosis physical activity was significantly associated with an improved cancer-specific survival [hazard ratio (HR) cancer-specific survival = 0.61, 95% confidence interval (CI) 0.44 to 0.86; random-effects model; p < 0.001]. The meta-analysis also found that higher post-diagnosis physical activity level was associated with a significantly increased overall survival (HR overall survival = 0.62, 95% CI 0.54 to 0.71; fixed-effects model; p < 0.001). Five31–35 out of the six individual studies assessing the relationship between physical activity level and overall survival found a statistically significant increase in overall survival among patients with higher post-diagnosis physical activity levels.

Pooled relative risks from another meta-analysis23 among CRC survivors showed inverse associations between post-diagnosis leisure-time physical activity and mortality based on six prospective cohorts. 30–35 The authors conducted a meta-analysis for exerciser versus not exerciser, moderate level of physical activity versus low physical activity, and high level of physical activity versus low physical activity. Low physical activity (reference group) was defined as 0, < 3 and < 3.5 MET hours per week or sedentary and the highest category was defined as ≥ 18 and ≥ 8.75 MET hours per week or sufficiently active. All categories above the reference group were pooled to represent ‘exercisers’. Exercisers had a risk ratio of 0.74 (95% CI 0.58 to 0.95; p = 0.02) for CRC-specific mortality, compared with non-exercisers. The risk ratios of CRC-specific mortality for moderate versus low physical activity and high versus low physical activity were 0.82 (95% CI 0.61 to 1.10; p = 0.19) and 0.65 (95% CI 0.47 to 0.92; p = 0.01), respectively. Similarly, exercisers had a risk ratio of 0.68 (95% CI 0.60 to 0.78; p < 0.001) for all-cause mortality, compared with non-exercisers, and the risk ratios of all-cause mortality for moderate versus low physical activity and high versus low physical activity were 0.76 (95% CI 0.64 to 0.90; p = 0.001) and 0.61 (95% CI 0.52 to 0.71; p < 0.001), respectively.

The meta-analysis indicated that there may be a threshold ‘dose’ of physical activity that is necessary to yield a health protective effect. 23 In this meta-analysis, those who participated in both moderate amounts of physical activity and high amounts of physical activity after diagnosis had a 24% and 39% risk reduction in all-cause mortality, respectively, compared with those who participated in low amounts of physical activity. Thus, the researchers highlight a clinically relevant finding from the meta-analysis, which is that a moderate amount of physical activity participation (defined by the researchers as physical activities between 3 and 18 MET hours per week or between 3 and 8.75 MET hours per week, or between 1 and 150 minutes of physical activity per week) was associated with a 24% risk reduction in all-cause mortality, whereas previously reported individual studies33,34 had suggested that higher levels of more than 18 or 27 MET hours per week of physical activity participation were associated with favourable survival outcomes in CRC survivors. 23

Another meta-analysis29 of the same studies30–35 also found a dose response. Each 5-, 10-, or 15-MET hours per week increase in post-diagnosis physical activity was associated with a 15% (95% CI 10% to 19%), 28% (95% CI 20% to 35%) and 38% (95% CI 28% to 47%) lower risk of total mortality, respectively. The reviewers also found that the apparent protection from total mortality afforded by physical activity was not modified by tumour stage, cancer treatment, smoking or adiposity.

A study published after the above meta-analysis36 found that spending ≥ 7 hours per week in leisure time physical activity after a diagnosis of CRC was associated with a 31% lower risk of death (from any cause) than doing no leisure time physical activity (HR = 0.69, 95% CI 0.49 to 0.98; p for trend = 0.01; adjusted for pre-diagnosis leisure-time physical activity). Looking at specific causes of death, spending ≥ 7 hours in leisure time physical activity after diagnosis, compared with none, probably reduces risk of death from CRC (HR 0.53, 95% CI 0.27 to 1.03; p for trend = 0.04) but not death from cardiovascular disease (HR 0.89, 95% CI 0.42 to 1.86; p for trend = 0.38).

Only one of the studies included in the above meta-analyses reported on sedentary time and mortality among people with CRC31 and found that sitting for ≥ 6 hours per day compared with sitting for < 3 hours per day after diagnosis was associated with a 27% increased mortality risk (95% CI 0.99 to 1.64). This specific study has added to the growing literature on the health risks of sedentary time, which is defined as waking activities performed in a seated or reclining posture that require very low energy expenditure (< 1.5 METs). 37–40 For this reason, our study is designed to assess the feasibility and acceptability of obtaining objective measures of both sedentary time and physical activity among people with CRC.

Another recent meta-analysis adopted a slightly different approach from the three described above. 41 Rather than focusing on assessing the amount or categories of physical activity at one point of time with cancer outcome, the focus was on the impact of actual changes of physical activity over time. Physical activity was required to be assessed at least twice, before and after diagnosis, or during two follow-up periods after diagnosis. Patients who increased their physical activity level or remained active throughout the diagnosis and treatment phases had significantly higher quality-of-life scores than patients with reduced physical activity levels after treatment. Pooled analysis from two studies30,35 found a significant association between increases in physical activity in the post-diagnosis period and reduced CRC death, but increases from pre diagnosis to 5 months post diagnosis were not significant. However, the overall pooled HR estimate of 0.70 (95% CI 0.55 to 0.85) indicated that the change was significant. A similar pattern was observed for overall mortality (pooled HR 0.75, 95% CI 0.62 to 0.87). Pooled analysis from three studies42–44 revealed that, compared with decreased physical activity post diagnosis and/or post treatment, increased physical activity was associated with significantly higher quality-of-life scores [standardised mean difference (SMD) 0.74, 95% CI 0.66 to 0.82]. The impact of physical activity on quality of life for people with CRC is discussed further in the following section.

Physical activity and psychosocial outcomes

Recent systematic reviews and meta-analyses of multiple cancer types provide evidence that physical activity interventions can help to address the psychosocial effects of cancer and associated treatments. 45–47 These reviews have been chosen for this report because they have been published recently (i.e. since 2010) and therefore include most recent evidence. Below we summarise the results of these reviews relating specifically to our outcomes of interest (i.e. level of physical activity, quality of life, fatigue, anxiety and depression). In addition, any results and conclusions about the moderating influence of intervention characteristics are described. Signs of effect sizes described below are set so that negative effect sizes for fatigue, anxiety and depression and positive effect sizes for quality of life and level of physical activity indicate improvements in favour of intervention participants.

A meta-analysis of 82 unique studies of interventions (not conducted in a physical therapy setting or delivered by a physical therapist) concluded that there is a small to moderate effect on physical activity level (weighted mean effect size 0.38, 95% CI 0.22 to 0.54; p = 0.0001), overall quality of life (weighted mean effect size 0.29, 95% CI 0.03 to 0.54; p = 0.03), and fatigue (weighted mean effect size –0.54, 95% CI –0.90 to 0.19; p = 0.003) in post-treatment interventions, and a small to moderate effect on functional quality of life (weighted mean effect size 0.28, 95% CI 0.02 to 0.54; p = 0.04) and anxiety (weighted mean effect size –0.21, 95% CI –0.39 to –0.03; p = 0.02) during treatment interventions. 45 No significant effect of physical activity interventions on depression was found (18 studies assessed depressive symptoms). Although intervention characteristics are described, no analyses of moderating effects are reported.

A meta-analysis of 15 RCTs of moderate-intensity physical activity programmes found a small but significant effect on depression under a random-effects model [effect size reported as mean change scores (Cohen’s d) –0.2, 95% CI –0.43 to –0.009; p = 0.04]. 46 Characteristics of physical activity interventions were found to be significant. Home-based exercise was associated with increased depressive symptoms [effect size reported as mean change scores (Cohen’s d) 0.16, 95% CI –0.15 to 0.47], compared with an improvement in depressive symptoms from interventions in other locations (e.g. community facilities, laboratories and gyms) [effect size reported as mean change scores (Cohen’s d) –0.45, 95% CI –0.77 to –0.14] and was significant (p = 0.04). Supervised and partially supervised exercise produced reductions in depressive symptoms, whereas non-supervised activity was associated with a small increase in depressive symptoms [supervised: effect size reported as mean change scores (Cohen’s d) –0.67, 95% CI –1.11 to –0.23; mixed supervision: effect size reported as mean change scores (Cohen’s d) –0.32, 95% CI –0.50 to –0.14; and unsupervised: effect size reported as mean change scores (Cohen’s d) 0.25, 95% CI –0.01 to 0.50] and was significant (p = 0.01). Exercise bout durations of > 30 minutes had larger effects on depression than exercise bouts of ≤ 30 minutes [> 30 minutes’ bout: effect size reported as mean change scores (Cohen’s d) –0.57, 95% CI –0.91 to –0.23; ≤ 30 minutes’ bout: effect size reported as mean change scores (Cohen’s d) 0.01, 95% CI –0.20 to 0.22] and this was significant (p = 0.02).

A meta-analysis of 34 studies of physical activity interventions for patients after they had completed their main treatment (it was possible that patients were still undergoing hormonal treatment) found effects on fatigue, depression and quality of life. 47 Measured by the revised Piper Fatigue Scale, physical activity was associated with slightly reduced fatigue (mean difference −1.0, 95% CI −1.8 to −0.1; p = 0.03) in three comparisons from two studies on breast cancer, compared with the control. In survivors of mixed types of cancer, physical activity was associated with reduced depression (mean difference −4.1, 95% CI −6.5 to −1.8; p < 0.01) as measured by the Beck Depression Inventory. In survivors of mixed types of cancer, physical activity improved the Short Form Health Survey 36 items (SF-36) physical function (mean difference 3.0, 95% CI 0.7 to 5.3; p = 0.01), social function (mean difference 3.4, 95% CI 0.4 to 6.4; p = 0.03) and mental health scores (mean difference 2.4, 95% CI 0.7 to 4.1; p = 0.01) compared with the control group. Although intervention characteristics were described, no analyses of moderating effects were reported.

These meta-analyses suggest that physical activity interventions have a small to moderate effect for level of physical activity, quality of life, fatigue, anxiety and depression. Although these reviews demonstrate patient benefit, a note of caution is required because most of the studies included in these reviews involve people diagnosed with breast cancer (e.g. 83% of all studies in the review by Speck et al. 45 involved people with breast cancer) and the results of these studies cannot be automatically generalised to people with CRC. This is because people with CRC, relative to people with a diagnosis of breast cancer (i.e. the patient group most represented in the controlled trials of physical activity), are likely to present with more advanced disease, have different treatments and side effects, tend to be older and include equal numbers of men and women.

However, a 2014 systematic review and meta-analysis of physical activity interventions for people with CRC suggests that, with appropriate support, people with CRC can increase physical activity levels, with subsequent health-related benefits. 24 Only three RCTs were included in the meta-analysis:44,48,49 a pilot RCT involving 18 participants who had recently undergone surgery and completed chemotherapy treatment for CRC, randomised to either a 12-week programme of twice-weekly supervised exercise sessions for 6 weeks followed by 6 weeks’ home-based exercise and dietary advice or standard treatment;48 a RCT involving 102 CRC survivors randomised to either an exercise group where they were specifically advised to perform moderate-intensity exercise 3–5 times per week or a control group who were requested not to exercise;44 and a RCT involving 46 people who had completed treatment for CRC and who were randomised to either 3 months of telephone counselling to support home-based physical activity or a contact telephone call. 49 The meta-analysis found no evidence for effects on quality of life (standardised mean difference 0.18, 95% CI –0.39 to 0.76; p = 0.53) or fatigue (standardised mean difference 0.18, 95% CI –0.22 to 0.59; p = 0.26). There was, however, evidence for improvements in physical fitness (standardised mean difference 0.59, 95% CI 0.25 to 0.93; p < 0.01). Analysis of moderating effects of intervention characteristics was not possible. The authors acknowledged that the review comprised only three studies with a total sample size of 166 participants and therefore concluded that further studies on physical activity for people with CRC are warranted.

At the time this current study was being developed (2012), two large-scale trials featuring people diagnosed with CRC were in progress. 50,51 The multinational, multicentre Colon Health and Life-Long Exercise Change trial (CHALLENGE) will determine the effects of a 3-year structured physical activity intervention on disease-free survival in 962 survivors of stage II or III colon cancer. 50 The 3-year intervention, modelled on the theory of planned behaviour, will consist of a behavioural support programme (n = 48 sessions in total), focusing on strategies to promote the adoption and long-term maintenance of physical activity and supervised physical activity sessions (n = 48 in total, one to one or in group format), designed to address physical activity techniques and intensity and safety, both delivered by a physical activity consultant (professional discipline not stated). The CanChange trial, drawing theoretically on cognitive–behavioural approaches for telephone-based health coaching sessions, recently reported significant intervention effects for moderate physical activity (28.5 minutes per week; p = 0.023). 51

What kind of cancer rehabilitation could be delivered in the NHS?

Given the increasing number of studies showing the safety and benefits of physical activity, it should be part of standard care for all cancer survivors. One barrier to physical activity becoming standard care is difficulties around implementation. Thus, when designing this study, we paid particular attention to this issue. We wanted a physical activity intervention that was effective, sustainable, cost-effective and capable of being integrated into the routine care of cancer survivors. As far as we know, this study is novel in that it aims to test an existing, evidence-based and theory-driven cardiac rehabilitation service for people with CRC. Should this model of rehabilitation prove to be clinically effective and cost-effective in a large-scale definitive trial, referral pathways could be adapted to ensure that the model is integrated into existing cancer service frameworks.

One of the reasons cardiac rehabilitation may be a suitable model for CRC survivors is that cardiac rehabilitation multiprofessional teams have the expertise required to provide relevant rehabilitation, including monitored physical activity, to a wide variety of patients, such as those with a CRC diagnosis. Second, a comparison of studies of coronary heart disease (CHD) and the post-treatment needs of people with CRC suggests that there is reasonable justification for referring CRC patients to cardiac rehabilitation and running mixed classes involving people with CHD and people with cancer. Four qualitative studies of patients’ experiences of needs after coronary artery bypass graft (CABG),52–55 one case note review of needs of 521 patients surgically treated for CRC cancer56 and one population-based cohort study including 522 people with CRC57 all indicate that people with CHD and people diagnosed with CRC experience similar problems, including pain, fatigue, anxiety and depression, worry, appetite loss, sexual problems, sleep disturbance, and work and financial-related difficulties, and express a need for information about medication and self-management. Thus, the rehabilitation needs of people with CHD and those of people with CRC are likely to be similar, suggesting that a common rehabilitation programme may be appropriate. Moreover, cardiac rehabilitation may be particularly relevant for people with CRC because the estimated prevalence of cardiovascular disease is 59% at 5 months post diagnosis, and, 16% develop de novo cardiovascular disease within 36 months of completion of treatment. 58 In addition, common comorbid conditions in CRC survivors include congestive heart failure, diabetes mellitus and chronic obstructive pulmonary disease,59 which, again, may be managed by rehabilitation.

Pointing out the similarities in post-treatment experiences is not to deny that there are, of course, disease-related differences among different patient groups. For example, people diagnosed with, and treated for, CRC can experience physical discomfort, bowel function problems and urinary tract infections and need advice about abdominal pain and stoma care,9 problems almost certainly not experienced by those with CHD unless they have comorbidities. There is a need, however, to meet the long-term rehabilitation needs of cancer survivors as well as a need to identify cost-effective and sustainable models of rehabilitation. This is why researching the effectiveness of rehabilitation to mixed classes of people with CHD and people with CRC is justifiable.

A further reason why it is worthwhile finding out if cardiac rehabilitation is a feasible, acceptable and effective model for CRC survivors is that cardiac rehabilitation is standard clinical care for patients after a cardiac event and is widely available throughout the UK. Comprehensive cardiac rehabilitation consists of supervised exercise, behavioural change, lifestyle risk factor management, education and psychological support. 60,61 In 2010–11 there were 276 cardiac rehabilitation centres in England, Northern Ireland and Wales62 and 37 in Scotland. 63

Cardiac rehabilitation is defined as:

The coordinated sum of activities required to influence favourably the underlying cause of cardiovascular disease, as well as to provide the best possible physical, mental and social conditions, so that the patients may, by their own efforts, preserve or resume optimal functioning in their community and through improved health behaviour, slow or reverse progression of disease. 61

Traditionally, the provision of cardiac rehabilitation has been described using phases 1–4, as mentioned in the National Service Framework for CHD. 64 Figure 1 summarises each phase.

FIGURE 1.

Four phases of cardiac rehabilitation.

A more recent approach in the Department of Health’s commissioning pack on cardiac rehabilitation65 describes cardiac rehabilitation along a best practice care pathway, using stages 0–6 to reflect core stages in the cardiac rehabilitation pathway, as shown in Box 3.

Another reason why cardiac rehabilitation may be a good model for cancer rehabilitation is that supervised exercise is the cornerstone of cardiac rehabilitation. 60,61 Although phase 3 cardiac rehabilitation varies across the UK, exercise classes are usually offered to patients once-weekly for 6–10 weeks. There are 13 rehabilitation standards for cardiac rehabilitation physical activity and exercise. 66 These provide a benchmark for all cardiac rehabilitation programmes delivered throughout the UK and include standards about initial screening and assessment, goal planning, exercise programmes and health and safety. A recent audit found that at 12 months after participation in cardiac rehabilitation there was a 14 percentage point increase in the number of people exercising five or more times a week for 30 minutes and a 23 percentage point reduction in those who rarely/never took exercise. 62 These changes in levels of physical activity represent important milestones for achieving recommended physical activity levels associated with disease prevention. 26 There is no obvious reason why the observed increases in the amount of physical activity among people with CHD could not also be found among people with cancer attending cardiac rehabilitation.

Until the publication in 2012 of a study contesting the benefits of cardiac rehabilitation,67 the consensus had been that cardiac rehabilitation decreases mortality rates in people with CHD. After we had started our study, however, the Rehabilitation After Myocardial Infarction Trial67 involving 1813 patients was published, which reported that there were no significant differences between patients referred to rehabilitation and controls in mortality at 2 years (risk ratio 0.98, 95% CI 0.74 to 1.30) or after 7–9 years (0.99, 95% CI 0.85 to 1.15), cardiac events, quality of life or psychological general well-being. 67 Its publication has caused considerable debate in those providing cardiac rehabilitation services,61,68 not least as the results seem to contradict a 2011 Cochrane systematic review and meta-analysis, which included 47 studies with over 10,000 patients and showed that cardiac rehabilitation reduces death from any cause by 13% and cardiac deaths by 26%. 69 Our study, therefore, took place at a time when there was a degree of controversy about the health benefits of cardiac rehabilitation for people with CHD.

The importance of feasibility and pilot studies

Conducting a full-scale RCT and economic evaluation of cardiac rehabilitation versus usual care for CRC survivors requires the involvement of many sites and is likely to be resource intensive. As there are uncertainties regarding rates of eligibility, consent, recruitment, retention and participation in the intervention and uncertainties about the feasibility and acceptability of the intervention for patients and clinicians, it is important in the first instance to conduct feasibility and pilot work. Findings from feasibility and pilot work can then be used to optimise the design and conduct of any subsequent large-scale trial or, indeed, be used to judge whether or not it is even appropriate and ethical to proceed to such a trial.

Conducting pilot studies to iron out methodological bias in advance of a large-scale trial is critical if that larger trial is to become part of an evidence base that is then used for recommending policy and changing cancer care practice. The importance of addressing methodological bias was highlighted in a recent meta-analysis of 33 RCTs of physical activity interventions for people with breast cancer. 70 The meta-analysis found that RCTs rated at high risk of selection bias with the absence of random sequence generation, or at high risk of attrition bias with large attrition rate or the absence of intention-to-treat analysis, resulted in greater efficacy of physical activity on quality of life, anxiety or depression at the end of intervention in experimental group versus control. The reviewers call for exacting methodological standards in future trials to increase confidence in evidence about the benefits of physical activity for cancer survivors. A pilot trial in advance of a larger-scale trial may be one useful approach to improve methodological rigor and standards.

The Medical Research Council’s (MRC) recommended framework for the evaluation of complex interventions includes feasibility and piloting phases. 71 Currently, however, there are no internationally agreed definitions of feasibility and pilot work. 72 One definition of feasibility study and its differentiation from pilot study comes from the UK’s National Institute for Health Research (NIHR) Evaluation, Trials and Studies Coordination Centre:73

Feasibility Studies are pieces of research done before a main study in order to answer the question ‘Can this study be done?’ They are used to estimate important parameters that are needed to design the main study.

Pilot studies are a version of the main study that is run in miniature to test whether the components of the main study can all work together. It is focused on the processes of the main study, for example to ensure recruitment, randomisation, treatment, and follow-up assessments all run smoothly. It will therefore resemble the main study in many respects, including an assessment of the primary outcome.

Reproduced with permission from NIHR

Thus, whereas a feasibility study may examine a specific part of a trial, a pilot study is a dummy run, examining the trial as a whole in order to see if all of the parts work together as planned.

In a review of 54 pilot and feasibility studies,74 researchers found that pilot studies tend to have more rigorous methodological components, such as sample size estimation, randomisation and control group selection, and are more likely include a greater number of methodological components for testing than studies labelled as ‘feasibility’. Nevertheless, the reviewers drew the conclusion that the distinction between the two is not clear-cut. 74

Irrespective of what researchers actually call preliminary studies, it is important that they clarify why pilot and feasibility work is being carried out. Feasibility and pilot work can be conducted to evaluate the operational feasibility and acceptability of the intervention itself and the feasibility and acceptability of a trial’s protocol design. There seems little point in running large-scale (and therefore presumably expensive) trials of interventions – even those suggesting promise of effect – if these interventions are unlikely to ever see the light of day and be implemented in practice. Likewise, if a trial is unworkable, then results about effectiveness will not be forthcoming. Thus, for our study we explored the twin pillars of feasibility and pilot work by examining intervention implementation and trial methodology parameters.

Bowen et al. 75 recommend eight areas of focus to assess if a public health intervention is feasible (Table 1). Addressing each area can help in the assessment of the likelihood of an intervention being implemented as part of routine health care and as a future commissioned service. Their recommendations for areas of focus in feasibility studies share similarities with frameworks designed to identify public health impact of health promotion interventions and evaluate the extent to which it is implementable, such as RE-AIM. 76 Using this approach, feasibility and pilot work can be carried out to provide information that can be used to modify an intervention to enhance its future implementation, as well as inform decisions about whether or not it is sensible, from an implementation perspective, to progress to a large-scale trial.

Another reason for conducting feasibility and pilot work is to evaluate trial methodology. Thabane et al. 77 propose four primary purposes for conducting pilot studies (Table 2). Although these key reasons were initially identified to guide the conduct of drug pilot trials, they have recently been adapted and used to guide the conduct of a rehabilitation intervention pilot trial. 78 Addressing these four areas will give some indication of the chances of a large-scale trial being successfully conducted.

It is recommended that threshold criteria for claiming future success of a large-scale trial are established before feasibility and pilot work commences. 77,78 The criteria should be based on the primary feasibility objectives. 77 Examples include the acceptable proportion of participants being eligible, consenting and completing the intervention. Using these criteria, the outcome of a pilot study will be one of the following:

1. stop – main study not feasible

2. continue, but modify protocol – feasible with modifications

3. continue without modifications, but monitor closely – feasible with close monitoring

4. continue without modifications – feasible as is.

Using this approach, pilot and feasibility studies provide critical information for planning and designing large-scale trials and justification for whether or not to allocate large sums of money to such a trial.

It is generally recommended that feasibility and pilot studies descriptively evaluate a trial’s feasibility, acceptability and safety rather than test the effectiveness of the hypotheses of the planned main large-scale trial. 74,77,79,80 This is because the small number of effect data available in feasibility and pilot studies mean the degree of uncertainty is such that the chance of reaching inaccurate conclusions about intervention effect is high. Feasibility and acceptability assessments of trial components may also be misleading if only a limited number of highly motivated sites are included in a pilot study because these sites are unlikely to be representative of the multitude of sites involved in a large-scale trial. 80 Event rates such as recruitment and willingness to be randomised cannot be accurately estimated from small pilots and estimates of variance of the outcome variable to calculate sample size from small pilot studies are also likely to suffer from imprecision. 78–80 How many total participants are required to estimate a standard deviation (SD) for a sample size calculation is unclear, with suggestions ranging from 24,81 to 30,82 to 5083 and to 70. 84

Good trial design requires the magnitude of the clinically important effect size to be stated in advance, and at least some indication of the efficacy of the proposed intervention is often required to justify to funders that it is worth the effort and expense in conducting a large-scale trial. One strategy for reporting outcomes from pilot work is to declare ‘potential efficacy’ if the CI around the estimated effect of the intervention on a clinically important outcome includes a predefined minimal important difference and, conversely, to declare ‘potential harm’ if the harm effect lies outside the upper confidence limit for safety. 79 This approach acknowledges the limited power of pilot trials to confirm the benefits and/or harms of treatment, while at the same time minimises the likelihood of the abandonment of a large-scale trial on the basis of negative or positive results. 74 Nevertheless, conducting analyses to glean information about efficacy from pilot trials, although tempting, is misleading and unreliable. 76,78 As a consequence, any observed potential patient benefit ought to be reported extremely cautiously or not at all; robust and rigorous assessment of an intervention’s therapeutic implications must await adequately sized definitive pivotal trials. 80

Study aims

The CRIB (Cardiac Rehabilitation In Bowel cancer) study was funded by the NIHR Health Services and Delivery Research programme. The overall aims of the CRIB study were to assess whether or not using phase 3 cardiac rehabilitation is a feasible and acceptable model of rehabilitation to aid the recovery of CRC survivors (i.e. examine intervention implementation potential) and to test the feasibility and acceptability of the protocol design (i.e. examine methodological standard). Thus, the overall purpose of the study was to assess whether or not it is appropriate to progress to a larger-scale trial and, if so, to optimise the design and conduct of any such trial.

Study design

The CRIB study was set up to evaluate the feasibility and acceptability of an innovative approach to aid the post-treatment recovery of people with CRC available in the NHS. Specifically, this involved an intervention study to test the feasibility and acceptability of the referral of people who had recently had surgery for CRC, who may or may not have been receiving adjuvant therapy, to cardiac rehabilitation. We undertook a phased programme of work comprising intervention testing and feasibility work (phase 1) and a pilot RCT (phase 2). The pilot trial was supplemented by a preliminary economic evaluation to consider the cost-effectiveness of providing the intervention compared with usual care. There was also a qualitative component to explore the views and experiences of patients and clinicians involved in the study. A description of the study protocol has already been published. 85

In phase 1, we sought to answer the following research questions:

1. What modifications, if any, are required to be made to existing cardiac rehabilitation (the intervention) to make it more relevant and acceptable to CRC patients and clinicians?

2. What modifications, if any, are required to be made to the training and support provided by the cancer-exercise specialist to make it more relevant and acceptable to cardiac physiotherapists running the cardiac rehabilitation exercise classes?

3. What modifications, if any, are required to be made to the proposed trial procedures to make the trial more feasible to conduct and to make the trial procedures more acceptable to CRC patients and clinicians?

In phase 2, we sought to answer the following research questions:

1. Are participating centres likely to recruit a sufficient number of patients to deliver a large-scale trial?

2. What are the likely eligibility, consent, recruitment, adherence and completion rates and speed of recruitment for a future large-scale trial and how can these be optimised?

3. Are patients allocated to the control group also increasing levels of physical activity (i.e. contamination)?

4. What are the likely completion rates at baseline and follow-up for the proposed outcome and process measures for a future large-scale trial and how can these be optimised?

5. What sample size is required to power a future large-scale trial?

6. Have practitioners delivering the intervention delivered it as intended and in accordance with the study protocol and how can intervention fidelity be optimised for a future large-scale trial?

7. What are the enablers and barriers that clinicians experience in delivering rehabilitation for patients and in conducting the trial?

8. What are the enablers and barriers that cancer patients experience in participating in the rehabilitation programme?

9. Can costs and health and outcomes be measured for this group using the patient-reported outcome tools for use in the economic evaluation?

Objectives

Ethics approval and research governance

An application for NHS ethics approval for the CRIB study was submitted using the electronic Integrated Research Application System. A submission was made on 20 January 2013, received by the NHS ethics committee on 25 January 2013 and reviewed by the committee at a meeting on 14 February 2013 [Research Ethics Committee reference 13/NS/0004; Integrated Research Application System project identification (ID) 121757]. The committee requested further information and submission of revised documentation. Hence, revised documentation was submitted to the chairperson of the ethics committee on 21 February 2013 and a favourable ethics opinion was given on 22 February 2013.

Applications for NHS Research Management approval, an additional approval required in the UK for research involving NHS patients, staff or premises, were made to the research and development office in each of the three health boards in which the study was conducted. Approval was given on the following dates:

• site 1: 5 March 2013

• site 2: 17 December 2013

• site 3: 14 January 2014.

Trial registration

The trial was registered with the International Standard Randomised Controlled Trial Registry under the reference number ISRCTN63510637; and also with the UK Clinical Research Network Portal under the reference number 14092.

Phase 1 design

Phase 1 was a before-and-after study; this is a rigorous design in which dependent variables are measured before and after an intervention has been delivered. 86 This design is suitable for assessing the feasibility of delivering an intervention and main trial procedures.

Participants

Recruitment procedures

Recruitment took place over 5 months. The first participant was recruited on 12 August 2013 and the last participant was recruited on 26 November 2013.

The following recruitment procedures for phase 1 were employed.

A CRC clinical nurse specialist assessed all CRC patients admitted for surgery to determine their eligibility for the study. At a follow-up appointment, the nurse gave eligible patients an information sheet (see Appendix 1) about the study, talked them through it, and completed a screening and recruitment form (see Appendix 2) for all eligible patients. This form included, for instance, information about a patient’s demographic characteristics (e.g. age and gender), cancer diagnosis (e.g. rectal or colon), date and type of surgery (e.g. open surgery or laparoscopic), and neo-adjuvant and adjuvant therapies. The patient signed this form if they were willing to participate in the study and agreed to have their contact details forwarded to an investigator. If the patient agreed to participate, the nurse then referred the patient to cardiac rehabilitation by e-mail, fax or letter, and used a referral form (see Appendix 3) to advise cardiac rehabilitation services that the patient would be participating. Patients who, having read the study information, declined to participate were asked if they were willing to give their reasons for declining, which were recorded by the nurse on the ‘reasons for not participating’ form, and the patient was asked to sign a non-participation patient consent form (see Appendix 4) if they were willing to have information about them (e.g. age, gender, diagnosis, treatment) used by the investigators to assess if participants were representative of the study population.

An investigator contacted the patients who had signed the screening and recruitment form, agreeing to participate, and arranged a time for them to sign a consent form, which meant that they had formally consented, in writing, to participation in the study (see Appendix 5). Baseline assessments were conducted for consenting participants.

Once a CRC patient had consented to the study, a member of the cardiac multidisciplinary team (e.g. cardiac physiotherapist or nurse) contacted them and invited them to attend a cardiac rehabilitation clinical/risk stratification assessment to determine whether or not, from a cardiac clinical perspective, they were able to exercise safely; the team member also planned physical activity goals tailored to the individual patient’s needs. Patients deemed safe to exercise were invited to attend cardiac rehabilitation classes.

A flow chart outlining CRIB phase 1 recruitment and follow-up procedures is given in Figure 2.

FIGURE 2.

Flow chart outlining phase 1 recruitment and follow-up procedures. CR, cardiac rehabilitation.

Informed consent

Informed consent for patients with CRC was obtained at two stages of the recruitment process. First, nurses obtained written consent to forward a patient’s contact details to an investigator if trial eligibility was established. The patient signed a screening and recruitment form (see Appendix 2). Second, an investigator obtained written consent before undertaking the baseline assessment. The patient signed a consent form (see Appendix 5). The original signed and dated consent forms were held securely as part of the trial site file, with a copy in the clinical notes held securely at the hospital.

Thus, all participants gave written consent to participate in the study. Informed consent discussions for participants took place face to face with a nurse and an investigator, with the opportunity given for participants to ask questions. Patients were informed that they had no obligation to participate and their care would not be affected if they declined to participate. They were made aware that the results of the study would not directly give rise to changes in rehabilitation provision for CRC patients; rather, it would determine whether or not large-scale trials, which may give rise to change in rehabilitation, were feasible. If a patient’s consent to participate in the study was declined or terminated at any stage, that patient then entered usual follow-up care.

Participants had the right to withdraw from the study at any time for any reason, and without giving a reason. The investigator also had the right to withdraw patients from the study intervention if this was considered in the patient’s best interests. There were two withdrawal options:

1. complete withdrawal from both the study intervention (i.e. cardiac rehabilitation) and the provision of data

2. partial withdrawal, when the patient withdrew from participating in cardiac rehabilitation but continued to provide data.

Consent was sought from participants choosing option 1 to retain data collected up to the point of withdrawal. Participants were also asked if they would be willing to give their reasons for their decision to withdraw so that these could be recorded, as this would help to improve acceptability of the study in a large-scale trial. We also gathered data about patients with CRC who declined to participate in the study to explore their reasons for not giving consent, thereby helping us to make the study more acceptable to patients in a large-scale trial.

Intervention

Measures

The primary objective of a future large-scale RCT will be to test if cardiac rehabilitation is clinically beneficial for CRC survivors and cost-effective. As we have explained in Chapter 1, there is evidence that physical activity is associated with improved survival and quality of life, and with reduced anxiety, depression and fatigue. At the time of designing this study, the proposed primary outcome for a future large-scale trial would be the difference in measures of physical activity (e.g. minutes per week, MET hours per week, time spent sedentary and in moderate-intensity activity) between the intervention and usual care (control) groups, measured by accelerometer. The proposed secondary outcomes were self-reported measures of quality of life, anxiety, depression and fatigue. In phase 1, we assessed the feasibility and acceptability of data collection instruments for these proposed outcomes. The results of phase 1 informed decisions about which data collection instruments would be tested further in phase 2 (i.e. pilot RCT) or replaced. An economic evaluation was also planned for any future large-scale trial, and so questions for this economic evaluation were also assessed for feasibility and acceptability during phases 1 and 2 of this study.

The following accelerometer and patient-reported outcome measures were taken at baseline (T0) and approximately 2 weeks post intervention (T1). Information about the validity and scoring of each measure is provided below.

Sample size justification

The sample size was based on two factors: (1) number of patients with CRC that was sufficient to address phase 1 objectives, that is, test the feasibility and acceptability of the intervention and main trial components before commencing phase 2 (i.e. a pilot RCT); and (2) estimated number of patients with CRC who could be recruited within a planned recruitment period of 2 months (the recruitment period was extended from 2 to 5 months).

As highlighted in Chapter 1, there is no clear guidance for how many participants are necessary for estimating event rates such as recruitment and willingness to be randomised in pilot RCTs. Similarly, there is no clear guidance for how many participants are required to assess the feasibility and acceptability of an intervention and study instruments. Thus, we aimed to recruit 12 patients with CRC, as this is the number of patients we thought that we could realistically recruit within the given time scale to meet phase 1 objectives.

Data collection and management

The original plan, as specified in the protocol, was to ask patients to complete the self-report questionnaires using pen and ink in the presence of an investigator who would guide them through it. However, all questionnaires were administered online (Bristol Online Survey) to save the time and expense of entering data.

Analysis

Adverse events

Although this was not a clinical trial of an investigational medicinal product, the investigators adhered to Tayside Medical Science Centre standard operating procedure 11: ‘Identifying recording and reporting adverse events [AEs] for clinical trials of investigational medicinal products’. 132 The following serious adverse event (SAE) protocols were reported within 24 hours of the principal investigator or person delegated responsibility for recording SAEs becoming aware of them:

A SAE is any AE occurring that results in any of the following outcomes:

• death

• inpatient hospitalisation or prolongation of existing hospitalisation

• persistent or significant disability/incapacity.

The following protocol exclusions applied:

• hospitalisation for assault or accidental injury

• hospitalisation for pre-planned surgery.

The above protocol exclusions were recorded in the AE log (see Appendix 8) for the study and line listings were reported annually to ethics and the sponsor.

Each hospital, and hence each cardiac rehabilitation programme, also had a reporting system for AEs, and cardiac rehabilitation operates a system of incident reporting. Thus, AE reporting of study participants by cardiac rehabilitation were also recorded by an investigator.

Feasibility and acceptability of main trial components

Colorectal cancer patient recruitment rate

Figure 3 presents patient flow through the study. In total, there were 34 new patient admissions. A nurse gave 24 (70% of all patient admissions) eligible patients a study information sheet. Nurses completed a screening and recruitment form for 17 (71% of those receiving a study information sheet) of these eligible patients. Nurses did not complete a screening and recruitment form for seven eligible patients who had been given study information, because the patient was too unwell, had been discharged or had moved wards. Ten (58% of completed screening and recruitment forms) patients signed a screening and recruitment form indicating their willingness to participate in the study and to have contact details forwarded to an investigator. Six patients who signed a screening and recruitment form indicating their willingness to participate withdrew from the study before signing a consent form and entering the intervention owing to ill health (n = 3) or travel problems (n = 2) or because they were subsequently unable to be contacted (n = 1). Four of these patients signed a consent form and started cardiac rehabilitation (i.e. the intervention), which is 17% of eligible patients. One of these patients withdrew owing to ill health.

FIGURE 3.

Participant flow.

Feasibility and acceptability of the intervention

Patients’ and clinicians’ experiences and perceptions of cardiac rehabilitation

The themes were referral pathways to cardiac rehabilitation, importance of exercise for patients with CRC, cancer and cardiac patients exercising together, and cardiac rehabilitation education sessions.

Referral to cardiac rehabilitation

The cardiac rehabilitation physiotherapist did not find the referral procedures of CRC patients into the service acceptable and so these were changed. The original plan, as stated in the protocol85 and described in Chapter 3, was that nurses would refer patients to cardiac rehabilitation. In the absence of a study, this is how patients would be referred if the intervention were implemented as part of routine care. In this study, however, the investigator referred patients to cardiac rehabilitation. The reason for this was to minimise the workload of the cardiac rehabilitation team. If we had kept to the original procedure (i.e. nurses directly referring patients to cardiac rehabilitation), the onus would have been on the cardiac rehabilitation team to continuously check with the patient when they felt ready to attend cardiac rehabilitation. The difficulty was that patients with CRC varied in terms of their recovery and readiness to start cardiac rehabilitation following surgery, which meant that the time to starting cardiac rehabilitation after surgery could not be strictly regimented. To save the team repeatedly contacting CRC patients to find out if they were ready to start the programme, telephone calls by an investigator were introduced instead. It was only once a CRC patient informed an investigator that he or she was ready to attend cardiac rehabilitation that an investigator informed the cardiac rehabilitation team about that patient. A member of the cardiac rehabilitation team could then contact the patient to invite him or her to attend the first appointment to conduct a risk assessment and discuss the programme. A physiotherapist explained why the original procedure was problematic.

It was slightly cumbersome, I think with the to-ing and fro-ing between when we wanted to fit them in, checking with you [the researcher] when you were available [to consent the patient and collect baseline measures before the patient started cardiac rehabilitation] and regrouping, em, and also time wasted initially, em, but once we got round to you phoning them in advance it was fine.

CRP 001

The new procedure to be implemented for the purposes of the study was also endorsed by the focus group. In particular, the cardiac rehabilitation team was concerned about the amount of time required of them to include CRC patients in their service and so welcomed any procedure that would reduce this.

Nevertheless, all of the clinicians involved in phase 1 reported that direct referrals from the cancer care team to cardiac rehabilitation would be possible in the future if this model of rehabilitation were to be implemented as part of routine care. Figure 4 illustrates key differences between referral and enrolment procedures in the study and the procedures if this model of cardiac rehabilitation were to be rolled out as part of routine aftercare.

FIGURE 4.

Research and practice cardiac rehabilitation referral pathways. CR, cardiac rehabilitation.

The referral form sent to the cardiac rehabilitation team about a new patient was modified in the light of interviews and focus group discussions with clinicians. The additional information requested is given in Table 11.

TABLE 11 - Additional information requested

GP, general practitioner.

Information	Reason for request
Names and address of GP	This was included (with participants’ prior consent) in case it was felt necessary to inform the GP that a patient on their register was participating in cardiac rehabilitation
Date of surgery	This information was included because CRC surgeons suggested that it was safe to exercise 4–6 weeks post laparoscopic surgery and 6–8 weeks post open surgery. Thus, cardiac rehabilitation practitioners wished to know the date of surgery
Current medication	This was included for safety reasons in case the patient became unwell during exercise classes
Relevant past medical history	This was included for safety reasons so that cardiac physiotherapist could assess whether or not the patient was safe to attend exercise classes

Importance of exercise for patients with colorectal cancer

All three CRC nurses were supportive of cardiac rehabilitation for patients with CRC because it would help with patient recovery.

The more exercise you do post-op, you do, you have a much better recovery rate.

CNS 001

I like to try and encourage a healthy approach to living anyway so I think it [cardiac rehabilitation] formalises what I encourage.

CNS 002

Nurses mentioned the benefit of physical activity when recruiting patients to the study.

[The nurse said] it’s good for you, do it.

P005

Nevertheless, patients with CRC reported that they had not been informed about the role of physical activity to reduce the risk of cancer:

Yes. And are you aware of any things that you can do yourself to reduce the risk of the cancer coming back?

No, I haven’t been told about anything.

P005

And, how do you think you can reduce the risk of the cancer coming back, have you been given any steps that you can take?

No.

P006

Patients with CRC welcomed the opportunity to attend cardiac rehabilitation for three reasons: physical activity is generally beneficial, it is difficult to exercise independently, and rehabilitation provides an environment in which patients can learn how to exercise safely. These beliefs in the benefits of cardiac rehabilitation are likely to have influenced patients’ willingness to participate in the study.

Two patients with CRC said that being physically active would be good for them.

I’m sure that any exercise is good for us . . . em, sitting is maybe the worst thing you can be doing.

P006

You’ve got to be willing to try different things, it’s for your own good really, you want to try as much as you can.

P001

One patient with CRC said that she was not very good at exercising on her own at home.

. . . I mean, I do know it [exercise] does [help], but I’m not very good at doing it at home by myself [laughs].

Right, right, so you enjoy coming to a class to do it?

Eh, yes, it’s the getting here isn’t it but yes, it’s OK once I’m here it’s good for me.

P005

One patient with CRC welcomed the opportunity to attend cardiac rehabilitation because he learnt how to safely pace himself when being physically active.

It lets you know you can safely push yourself a bit.

P006

Cancer and cardiac patients exercising together

The intervention was referral of patients with CRC who had recovered from surgery to cardiac rehabilitation. Thus, as the physiotherapist pointed out, it is important to explain why this particular model for rehabilitation for patients with cancer was being researched.

Initially I just thought why involve them with cardiac rehab, why not set up a pilot geared towards cancer patients?

CRP 001

The physiotherapist found it straightforward to slot patients with CRC into existing classes.

They have come in as normal patients . . . they just slipped in and joined in.

CRP 001

Patients with CRC did not perceive mixing with cardiac patients during rehabilitation as problematic:

And how do you feel about it being a mixed group?

Oh, that didn’t bother me at all . . .

At all?

No, no.

No, OK.

I quite enjoy the various company of people, yes.

P005

And what about it being a group with cardiac patients as well?

You do talk to other people . . . most were heart, I think I was the only one with cancer . . . eh, maybe not.

Maybe, OK.

. . . Meeting other people and sitting having a cup of tea helps.

P001

Another patient summarised the mixed groups simply:

You just accept it, and they [cardiac patients] accept you.

P006

Cardiac rehabilitation clinicians attending the focus group also endorsed mixed patient exercise classes.

Cardiac rehabilitation education sessions

As described in Chapter 3, the intervention comprised a 60-minute exercise session followed by an education session. These education sessions were designed to provide information useful to cardiac patients. Nevertheless, according to the cardiac physiotherapist, patients with CRC attended not only the exercise sessions but also the education sessions.

And how do you feel that the patients that have come through have got on generally?

Well, all three of them have fitted in in fact very well, em, in fact, one of them, well they’ve just gone in to all the cardiac talks regardless.

One of the patients with CRC thought that the education sessions were the most useful part of the rehabilitation programme. This patient had recently had a stroke and so may have found the education sessions particularly relevant. However, given increasing comorbidity in older age, education sessions designed for cardiac patients may be relevant to patients with cancer.

. . . and the lecture is one of the best things about it.

Right, so what things, what particular lectures have you found, did you find most helpful?

Eh, drugs and eh, eh, resuscitation, although I knew something about resuscitation. But eh, eh, just em, the, letting you know about your condition and eh, eh, how to eh be a good boy and take care of it.

Not all education sessions, however, were regarded as useful. One patient with CRC said that the session on smoking cessation was not relevant because most patients, including cardiac patients, had given up smoking.

. . . I don’t know whether, it may have applied, most people had given up, there was only one person who was in the process of giving up and I really felt it wasn’t beneficial to anybody . . .

P005

As described in Chapter 3, it was planned that a CNS would deliver education sessions about cancer to patients with CRC. However, this did not happen. Thus, the cardiac rehabilitation programme was not altered in any way to accommodate CRC patients and patients with CRC attended classes alongside cardiac patients.

However, the cardiac physiotherapist believed that cancer education sessions should be provided:

I feel slightly for them that they’re getting all this cardiac stuff and not really any follow-up . . . there’s no support from the, from the cancer side as it was mooted about, eh, when we were talking about the project that, eh, if it was cancer-specific then they would maybe, eh, if it was cardiac-specific talks then they would maybe get some additional support and there’s just been absolutely nothing.

CRP 001

One patient also expressed a desire for cancer education:

. . . it’s more for heart than for cancer . . . I think that’s where it maybe falls down you know . . . but I don’t know how you will do that. When push comes to shove, they really didn’t know a lot about the cancer end of things, even [name of cardiac rehabilitation] said . . . eh . . . em, she didn’t know a lot about it [cancer].

P001

Summary of findings

The results for phase 1 are clear:

• Nurses were willing to recruit patients with CRC to the study. The evidence is that nurses gave a study information sheet to 70% (n = 24) of all surgical CRC patients and completed a screening and recruitment form for 71% (n = 17) of these patients.

• The majority of eligible patients were willing to participate in the study. The evidence is that 10 out of 17 eligible patients (58%) signed a screening and recruitment form indicating their willingness to participate in the study and have their contact details forwarded to an investigator.

• The surgical ward, when patients with CRC were admitted or while they were waiting to be discharged from hospital, was an appropriate time to raise the study with patients with CRC and to give them a study information sheet. The evidence for this is that 10 out of 17 eligible patients (58%) signed a screening and recruitment form indicating willingness to participate in the study before being discharged from hospital.

• There were barriers (e.g. travel and poor recovery from surgery) to patients participating in the intervention (i.e. cardiac rehabilitation), which had a detrimental impact on sample attrition. The evidence for this is that only 4 out of 10 (40%) patients who signed a screening and recruitment form then proceeded to start the intervention. These barriers and subsequent loss to the study of six participants are the main reasons why we did not meet our anticipated target of recruiting 12 patients.

• A good cross-section of patients with CRC were interested in taking part. The evidence is that men and women from the ages of 50 to 89 years, with and without metastatic disease, having open surgery and laparoscopic surgery, with and without a stoma, and having and not having adjuvant therapy were willing to participate in the study (i.e. signed a screening and recruitment form).

• An accelerometer was an acceptable objective method for assessing level of physical activity and sedentary behaviour. The evidence is that participants wore the accelerometer for a validated period of time.

• Investigator administration of online patient-reported outcomes questionnaires was an acceptable method for collecting outcomes data and was not perceived as a major burden by participants. The evidence is that all questions were answered at baseline and follow-up and there were no missing data.

• Some parts of the questionnaire were considered by participants as repetitive and irrelevant. Some participants found questions about quality of life very similar. Quality of life questions also failed to capture issues relating to the impact of CRC and treatments.

• A cancer and exercise expert delivering 1-day cancer and exercise training face to face or by video conference to cancer nurses and cardiac rehabilitation practitioners was feasible and acceptable. The evidence is that the results of the evaluation of the training were excellent with all attendees, for instance, either agreeing or strongly agreeing that the information and course content was helpful and well presented.

• Patient readiness to start cardiac rehabilitation varied owing to different rates of recovery and hence the length of time from date of surgery to the start of cardiac rehabilitation also varied (54–73 days).

• Adherence was suboptimal. The evidence is that participants were expected to attend 10 exercise classes but the four participants attended 10, 6, 5 and 0 classes, respectively.

• Cardiac rehabilitation for patients with CRC was feasible and acceptable. The evidence is that patients with CRC attended the exercise and cardiac-specific education sessions. However, it could be improved by introducing cancer-specific education sessions. In addition, cardiac and cancer clinicians were supportive of this model of rehabilitation and participated in the study.

Recommended changes to main trial components

Based on the results of phase 1, the following changes to the main trial components (Table 12) were recommended for further testing in phase 2.

Recommended changes to the cardiac rehabilitation

Based on the findings of phase 1, the following changes (Table 13) to the intervention were recommended for further testing in phase 2.

Limitations

Phase 1 was a small study. The feasibility and acceptability of trial components and the intervention were tested on only one site and included fewer than 10 participants. The site was a relatively small hospital serving a rural and geographically dispersed population. It is not possible to determine if the study is feasible and acceptable in different contexts, such as large urban hospitals with a relatively large number of trials simultaneously taking place. It is important, therefore, to test feasibility and acceptability in more sites with more participants.

Conclusions

The feasibility and acceptability of trial components and the intervention were tested on only one site over a short period of 6 months, involving a very small number of patients and clinicians. Phase 1 results suggested that nurses were willing to recruit patients with CRC to the study, the majority of eligible patients indicated a willingness to participate in the study, the surgical ward was an appropriate place for nurses to give study information, a good cross-section of patients with CRC were interested in taking part, and no data were missing from accelerometers or questionnaires. Phase 1 results also highlighted barriers to CRC patient participation, a suboptimal consent rate (we did not meet our anticipated target of recruiting 12 patients), repetition and lack of relevance in the self-report questionnaires, a long period of time between CRC patients indicating a willingness to participate and actually starting cardiac rehabilitation classes, suboptimal intervention adherence and lack of cancer-specific education sessions and lifestyle advice.

A decision was reached among the research team, funder and advisory group to proceed to phase 2 with the following main modifications to trial procedures and the intervention.

Phase 2 design

Phase 2 was a pilot RCT. As discussed in Chapter 1, and in accordance with the MRC’s framework for the evaluation of complex interventions,71 pilot work is valuable for helping to optimise study design and study procedures before proceeding to a large-scale trial. A RCT is also a robust method for reporting any preliminary effects of the intervention, with the proviso that these results are interpreted with great caution.

Participants

The study sought to recruit patients with CRC from two acute hospitals in Scotland and a large teaching hospital in Wales. Table 14 summarises characteristics for each hospital.

Recruitment procedures

Recruitment took place over 7 months. The first participant was recruited on 13 January 2014 (site 1) and the last participant was recruited on 29 July 2014 (site 2).

Based on phase 1 results (see Chapter 4), changes were made to the recruitment procedures and approved by the NHS Research and Ethics Committee (AM01 12/12/2013). The amended recruitment procedures in phase 2 were as follows.

A CRC clinical nurse specialist assessed patients admitted for surgery for CRC to determine their eligibility for the study. Eligible patients were given an information sheet (see Appendix 1) by the nurse (pre or post surgery) and the details of the study were discussed. If the patient agreed to participate in the study, the nurse asked them to sign a screening and recruitment form (see Appendix 2), before discharge from hospital, indicating their willingness and agreement to have their contact details given to an investigator. Patients who, having read the study information, declined to participate were asked if they would indicate their reasons for declining to participate on the screening and recruitment form; in addition, the patient was asked to sign a non-participation consent form (see Appendix 4) if they were willing to have information about them (e.g. age, gender, diagnosis or treatment) used by the investigators to assess if participants were representative of the study population.

After discharge from hospital, an investigator contacted patients who agreed to participate in the study, by telephone, to confirm the patient’s willingness to participate in the study and to establish if he or she was ready to attend cardiac rehabilitation. If the patient was willing and ready to attend cardiac rehabilitation, a mutually convenient time for the patient to meet with the investigator was arranged and formal written consent to participate in the study was sought. If the patient did not feel able to attend cardiac rehabilitation at this point in time (e.g. owing to poor recovery or transport difficulties) but was still willing to participate in the study and attend cardiac rehabilitation at some point in the future, then the investigator agreed to contact the patient again by telephone at a later date. Thus, at the first meeting with the investigator, eligibility was confirmed and written consent was obtained. If patients declined to give consent after hearing what the study involved, they were asked if they were willing to give their reason for no longer wishing to participate. Consented patients had baseline measures taken and were given an accelerometer to wear for a period of 7 days. They were given a FAQ (frequently asked questions) sheet, which listed answers to common queries (see Appendix 9). The patient was then randomised to the intervention or the control group (see Randomisation, concealment and blinding, which describes randomisation procedures).

Patients randomised to the control group were informed that they would not receive the intervention, but were given an information leaflet. They were also advised how and when to return their accelerometer to the investigator. Patients randomised to the intervention group were informed that they would be referred to cardiac rehabilitation.

The investigator completed a referral form and sent it on to the cardiac rehabilitation service. A member of the cardiac multidisciplinary team (e.g. a cardiac physiotherapist or nurse) then contacted the patient and invited them to attend a cardiac rehabilitation clinical/risk stratification assessment to determine whether or not the patient was able to safely exercise from a cardiac clinical perspective. Physical activity goals tailored to individual patient needs were also usually discussed at this time. Patients who were deemed safe to exercise were then given a date to start cardiac rehabilitation.

The recruitment process at site 3 was slightly different because a research nurse carried out all of the tasks conducted by the nurses and the investigator described above. Research nurses are nurses employed by hospitals in the UK to recruit to RCTs and can be working on a large number of trials simultaneously. Decisions about the use of research nurses for specific projects are decided by research and development managers in each NHS board. Site 3 had agreed that our study was able to make use of the research nurse, whereas the other two sites had not.

A flow chart outlining CRIB phase 2 recruitment and follow-up procedures is given in Figure 5.

FIGURE 5.

Flow chart outlining phase 2 recruitment and follow-up procedures. CR, cardiac rehabilitation.

Informed consent

The phase 2 informed consent procedures were identical to those used in phase 1 and are described in Chapter 3.

Randomisation, concealment and blinding

Eligible and consenting patients were individually randomised at the end of their baseline assessment to one of two treatment groups: cardiac rehabilitation (i.e. the intervention) or usual care plus a booklet with information about lifestyle following CRC diagnosis and treatment (see Treatment group allocation).

To conceal the allocation of treatment from those conducting the research, the randomisation of individual participants to a particular treatment arm was undertaken using an automated online randomisation system, which was administered remotely and used a computer-generated code. The randomisation service was provided by Tayside Clinical Trials Unit, a UK Clinical Research Collaboration-registered trials unit.

Once the randomisation procedure had been completed, the outcome and further details about the allocated treatment were immediately communicated by the investigator to the participant. Because of the nature of the intervention, it was not possible to blind participants, investigators or the clinicians delivering the intervention (i.e. cardiac rehabilitation multidisciplinary team) to the treatment allocation.

Randomisation was stratified to take account of the three sites. This was to ensure equal distribution of intervention and control participants per site.

Treatment group allocation

Measures

The primary objective of a future large-scale RCT will be to test if cardiac rehabilitation is clinically beneficial for CRC patients and also if it is cost-effective. As explained in Chapter 1, there is strong evidence that physical activity is associated with survival, improved quality of life and reduced levels of anxiety, depression and fatigue. At the time of designing this study, the proposed primary outcome for a future large-scale trial would be the difference in measures of physical activity (e.g. MET hours per week; sedentary and moderate activity) between the intervention and usual care (control) group measured by accelerometer. The proposed secondary outcomes in a future trial would be self-reported measures of quality of life, anxiety, depression and fatigue. An economic evaluation would also be conducted.

Based on phase 1 results, some data collection instruments were replaced for testing in phase 2 (see Chapter 4). All measures used in phase 2 are described below. A summary of outcomes and associated measures is presented in Table 18.

Sample size justification

The aim of the study was not to provide a definitive estimate of treatment effect, so we did not have a formal sample size calculation. Rather, the aim was to provide robust estimates of the likely rates of recruitment and retention, and to yield estimates of the variability of the primary and secondary outcomes to inform power calculations for a future large-scale effectiveness trial. As highlighted in Chapter 1, there is no clear guidance for how many participants are necessary for estimating event rates such as recruitment and willingness to be randomised in pilot RCTs. Our recruitment estimate for the pilot trial was based on three factors, which were the estimated:

1. number of patients admitted for surgery

2. number of patients likely to meet inclusion criteria

3. recruitment rates in previous similar studies (e.g. trials of physical activity interventions for people with cancer).

Based on information provided by the local NHS principal investigators of the number of patient admissions in the previous year (2012), we expected 250 patients, in total, to be admitted for surgery across the three sites over a 6-month period. Cancer clinicians involved in the study estimated that approximately one-third (n = 83) would be ineligible and, based on recruitment to a RCT about physical activity with patients with cancer in Scotland139 and a trial involving patients with CRC within 3 months of completing surgery conducted in Canada,44 we estimated that just over one-third of eligible patients would consent (n = 66) to take part. Thus, for the pilot RCT we estimated that we would recruit around 66 patients (40% of eligible patients). We estimated that sites 2 and 3 would each recruit 26 patients, and that site 1 would recruit 14 patients, as this site admitted fewer patients for surgery than the other two sites.

Data collection and management

Analysis

Phase 2 had seven objectives. To meet objective 1 (see Chapter 2), descriptive statistics were used to summarise screening, eligibility, consent, adherence and retention rates. Descriptive data on sample demographic characteristics were reported as frequencies, means and SDs. Differences in categorical variables of sociodemographic and clinical characteristics were compared between those who signed a screening and recruitment form but withdrew from the study and those who signed a consent form and were randomised to the intervention or control group. The reasons why eligible people with CRC did not wish to participate or withdrew, for the purpose of analysis, were sorted into one of seven categories:

1. no longer eligible

2. distance/travel

3. stated that they are currently exercising and fit

4. clinical (e.g. poor recovery from surgery, receiving adjuvant therapy, comorbidity)

5. too much of a commitment

6. not/no longer interested

7. no reason given.

To meet objective 2, completion rates and rates of missing data were assessed by counting the number of missed responses on the self-report questionnaires completed by participants at baseline and at 2- and 12-week follow-ups. In addition, the amount of time that participants wore the accelerometer at baseline and at 2- and 12-week follow-ups was counted (number of days worn over maximum wear-time of 7 days and number of hours per day worn).

To meet objectives 3 and 4, inferential statistics were used to assess any differences in outcomes between baseline and 2- and 12-week follow-up. These analyses are described in full in the statistical analysis plan (see Appendix 11). If data were normally distributed, outcome measures were assessed by multiple linear regression, adjusted for baseline variables; when data are measured more than once during the study, repeated measures analysis was also used. Scoring for outcomes follows the scoring instructions given for each questionnaire. When no such instruction is present, the following approach was taken: if no more than 20% of questionnaire items are missing, the missing items will be replaced by the mean of the remaining items to build a sum score. When more than 20% of the items are missing, the sum score will be set to missing. Outcomes were analysed as baseline versus end of intervention and baseline versus 3 months after intervention. An intention-to-treat approach was used for these preliminary analyses. Differences were considered statistically significant at p < 0.05.

To meet objective 5, descriptive statistics were generated of the proportion of CRC participants allocated to the intervention group who received cancer-specific education sessions or lifestyle advice from the CRC nurse specialists.

Objective 6 is addressed in Chapter 9, which describes the embedded qualitative study, and objective 7 is addressed in Chapter 10, which describes the economic evaluation.

Site recruitment

A total of six cardiac rehabilitation services and CRC clinicians located in the same hospital were initially approached to participate in the study. Two sites agreed to support the research. The reason why four sites did not wish to participate was due to cardiac rehabilitation concerns about capacity to include more patients in their classes. All CRC teams at the six sites were willing to collaborate in the study. Following contact with the NIHR Colorectal Cancer Clinical Studies Group, two further sites were identified and one of these sites was invited to participate in the study. Table 14 summarises the characteristic of the three sites involved in this study.

Flow of participants in the trial

In total, 41 individuals were recruited to the CRIB pilot trial, with 21 allocated to the intervention arm and 20 allocated to the usual care plus booklet group. Figure 6 presents the flow diagram for the trial and summarises patient throughput from referral to completion of the 1- and 3-month follow-ups. The diagram also reports the total numbers of patients who were given study information, did not meet inclusion criteria (sites 1 and 2 only), withdrew before randomisation, withdrew following randomisation or were lost to follow-up immediately following the intervention and at 3 months post intervention.

FIGURE 6.

Phase 2 total recruitment and sample attrition across all three sites.

Screening rate

Screening rate was defined as the number of people with CRC who were admitted for surgery and assessed for eligibility. Across all three sites, there were 198 people admitted to hospital for CRC surgery and, of these, 156 were assessed for eligibility (79%). Table 19 shows screening rates by site and shows that the research nurse in site 3 assessed 65% of patients for eligibility, whereas the clinical nurse specialists at the other two sites reached more patients, assessing 86% and 91% of all patient admissions, respectively.

Eligibility rate

Eligibility rate was calculated by dividing the number of people with CRC admitted for surgery by the number who met the inclusion criteria. Out of 198 patient admissions for CRC surgery, 133 (67%) met the eligibility criteria. Table 20 shows that the proportion of patient admissions for surgery who were eligible for the study was almost identical across the three sites (67%, 67% and 65%, respectively).

Consent rate

In total, 74 out of 133 eligible patients signed a screening and recruitment form indicating that they were interested in participating in the study and willing to be contacted by an investigator (i.e. 56% of the number of people with CRC who met inclusion criteria).

The consent rate was calculated by dividing the number of people with CRC who met the inclusion criteria (n = 133) by the number who consented to participate in the study and were randomised. Forty-one people with CRC consented and were randomised to the intervention or usual care (control) group, which is a consent rate of 31%. Twenty-one patients were randomised to the intervention group and 20 were randomised to the control group.

Table 22 shows the difference between total estimated and actual patient admissions, eligibility and randomisation rates across all three sites.

Figure 7 shows graphically the difference between the total estimated and actual patient admissions at each stage of the recruitment process. ‘Admissions’ refers to the numbers of estimated and actual patients admitted to hospital for CRC surgery; ‘assessed for eligibility’ refers to the number of these patients who were assessed for eligibility for the study; ‘eligible’ refers to the number of patients who met eligibility criteria; ‘screened’ refers to the number of eligible patients who completed a screening and recruitment form; ‘consent to approach’ refers to the number of patients who indicated on this screening form that they were interested in participating in the study and willing to have an investigator contact them about the study; and ‘randomisation’ refers to the number of patients who signed a consent form and were randomised to the intervention group or the control group.

FIGURE 7.

Total differences between estimated and actual admission, eligibility and randomisation rates.

The number of actual surgical admissions was lower than the protocol estimate (198 vs. 250). The actual number of eligible patients was approximately two-thirds of surgical admissions, which was on a par with the protocol estimate. In total, 133 out of 198 (67%) actual patient admissions were judged as eligible for the study. Seventy-four eligible patients signed a screening and recruitment form indicating willingness to participate and 41 of these patients consented and were randomised to the intervention or control groups. Thus, 31%, as opposed to an estimated 40%, of eligible patients consented and were randomised.

Retention rate

The retention rate was defined as the number of participants who remained in the study, that is, the number of participants who did not formally drop out. As indicated above, 41 people with CRC consented and were therefore identified as study participants. Three (7%) of these participants formally left the study (two control and one intervention).

Completion rates and missing data

Completion rate was defined as the number of consenting and randomised participants (n = 41) who completed outcome measures, that is, the number of participants who completed the self-reported questionnaires and were given an accelerometer to wear. Missing data were defined as the number of participants who were given an accelerometer but did not provide validated accelerometer data (see Chapter 6 for definition of validity) at baseline and at first and second follow-up.

Number of participants providing valid accelerometer data

Twenty-eight (68%) (14 intervention and 14 control) out of 41 participants provided validated accelerometer data to measure physical activity and sedentary behaviour at baseline and 23 (11 intervention and 12 control) (56%) and 14 (six intervention and eight control) (34%) participants provided validated accelerometer data at the follow-up 1 and follow-up 2 time points, respectively. These figures are broken down by site in Table 24.

TABLE 24 - Variation in complete data, by site

Site	Baseline, n (%)	Follow-up 1, n (%)	Follow-up 2, n (%)
Site 1 (n = 13)	9 (69)	9 (69)	8 (62)
Site 2 (n = 18)	11 (61)	7 (39)	2 (11)
Site 3 (n = 10)	8 (80)	7 (70)	4 (40)
Total	28	23	14

If all 41 participants had provided valid accelerometer data at each time point, there would have been a total of 123 accelerometer data sets. Sixty-five (53%) accelerometer data sets were provided and, of these, 20 (31%) were removed from analysis because data were invalid, which involved 14 different participants spread more or less evenly across the intervention and control groups. Table 25 shows reasons for invalid accelerometer data. The main reasons for invalidity were not wearing the device and not wearing it for a sufficient number of hours per day to meet the validation criteria.

TABLE 25 - Reasons for accelerometer data removal from analysis (N = 20)

Reasons invalid	Intervention, n (n = 11)	Control, n (n = 9)	Total, n (%)
Days worn (< 4)	0	3	3 (15)
Hours per day (< 10)	2	3	5 (25)
Not worn at all	5	2	7 (35)
Abnormal activity patterns	4	1	5 (25)

Figure 8 shows the number of completed questionnaires and valid accelerometer data at each time point.

FIGURE 8.

Number of participants completing self-reported questionnaires and valid accelerometer data.

Adverse events

In phase 2, no AEs were reported. One participant was unable to start cardiac rehabilitation owing to a torn knee ligament that was unrelated to the study.

Intervention adherence

As described in Chapter 6, adherence data were collected from the cardiac rehabilitation register of attendance and by a weekly telephone call to the participant. Adherence rate was calculated in two ways:

1. attendance during the cardiac rehabilitation programme (i.e. site 1: once-weekly for 10 weeks = 10, site 2: twice-weekly for 12 weeks = 24, site 3: twice-weekly for 6 weeks = 12)

2. attendance from start and end date that participant attended cardiac rehabilitation, including breaks and absences owing to, for instance, treatment.

Thirteen out of 21 participants (62%) randomised to the intervention group completed the programme as per protocol. Three participants started but could not complete all classes and five did not begin (38%). Table 26 shows that the main barrier to starting or dropping out of cardiac rehabilitation was poor physical health. Table 27 shows that participants who were able to continue with the programme had high levels of attendance (range 75–142%).

Participant length of time in the study

Table 28 shows baseline, follow-up 1 and follow-up 2 dates for participants who completed all time points (intervention, n = 13; control, n = 12). The mean number of days between baseline and follow-up was 86 and 104 for control and intervention group, respectively; between follow-ups 1 and 2 it was 83 and 78 days, respectively. The total mean number of days involved in the study was 169 and 182 for control and intervention groups, respectively. Thus, there was minimal difference in the number of days between data collection points between the two groups.

Comparability of characteristics of consenting and non-consenting eligible patients

Fifty-four out of 133 eligible patients who did not consent to participate in the study consented to have their demographic and clinical information used for the purposes of this study. Table 29 shows the characteristics of eligible consenting and not consenting patients. It shows no significant differences in age, gender and type of surgery (colon or rectal) between the two groups, but suggests that people with metastatic disease (T4 and N1 classification), people who have had open surgery and people with a stoma are more likely not to participate.

Baseline comparability of demographic and clinical characteristics of intervention and control group participants

Baseline physical activity self-efficacy and risk perception

A higher score indicates a higher level of physical activity self-efficacy (range 0–120). Table 33 shows that physical activity self-efficacy was high and skewed towards higher scores; at baseline the mean physical activity self-efficacy was 94.1 and 93.8 for intervention group and control group participants, respectively. A higher cognitive and affective risk perception score (range 0–10) indicates a greater perception that lifestyle behaviours have a protective health effect and a higher severity risk perception score (range 0–10) indicates a greater perception that CRC is a more serious disease. Risk perception scores were skewed towards higher scores. Table 33 shows little variation between the intervention and control groups; the table shows a mean cognitive risk perception score of 7.7 and 6.8, a mean affective risk perception score of 7.7 and 7.2 and a mean severity risk perception score of 6.1 and 5.6 at baseline for intervention and control group participants, respectively.

Physical activity and sedentary behaviour

As discussed in Chapter 1, results about the effects of an intervention measured in a pilot study should not be reported at all or reported only with great caution. As highlighted above, the completion rates for accelerometer measurement were low. We therefore present descriptive data only for the primary outcome, physical activity, and for sedentary behaviour; descriptive data for quality of life are presented in Appendix 12. Inferential statistical analyses are not reported.

Intervention fidelity

All CRC participants allocated to the intervention group reported that they had not received any additional educational sessions about cancer or lifestyle advice from CRC nurse specialists during the intervention.

Summary

Phase 2 shows that some of the key trial procedures were feasible and acceptable. The evidence for this is as follows:

• Colorectal cancer nurses screened 79% (n = 156) of all surgical CRC patients for eligibility.

• Sixty-seven per cent (133 out of 198) of all patient admissions for CRC surgery met eligibility criteria.

• Fifty-six per cent (74 out of 133) of eligible patients signed a screening and recruitment form indicating their willingness to participate in the study and to have their contact details forwarded to an investigator; 31% (41) were consented and randomised to the intervention or control groups. This was short of our target of 66 patients.

• A good cross-section of people with CRC were recruited. Men (n = 27; 65.9%) and women from the ages of 42 to 86 years (mean 66 years, SD 11.31 years), diagnosed with rectal (n = 16; 39%) or colon cancer (n = 25; 61%), having open surgery (n = 23, 56.1%) or laparoscopic surgery (n = 9; 22%), with (n = 13; 32%) and without a stoma, participated in the study.

• Only 7% (2 out of 41) of participants formally dropped out of the study.

• Forty (97.5%) out of 41 participants (20 intervention and 20 control) completed the questionnaires (SPAQ, FACT-C, EQ-5D, FACIT Fatigue Scale and HADS) at baseline, and 31 (75.6%) (15 intervention and 16 control) and 25 (61%) (12 intervention and 13 control) completed the questionnaires at follow-up 1 and follow-up 2 time points, respectively.

• No SAEs were reported.

• There was an insignificant difference in the number of days between data collection time points for participants in the intervention and control groups, and the mean numbers of days in the study were 182 and 169, respectively.

Phase 2 also shows potential threats to the internal and external validity of any future trial. In this chapter, we therefore make a number of recommendations to manage these threats. The evidence for potential risks to a future trial is as follows:

• Thirty-three out of 74 (44.5%) of eligible participants who signed a screening and recruitment form indicating willingness to participate in the study withdrew before randomisation. The main barrier to going on to actually participate in the study was poor health (e.g. poor recovery from surgery).

• Thirteen out of 21 participants (62%) who were randomised to the intervention group completed the programme. The main barrier to starting cardiac rehabilitation or stopping once starting to attend was poor physical health.

• Completion rates for accelerometer devices decreased over time. Sixty-eight per cent (n = 28) of participants provided validated accelerometer data to measure physical activity and sedentary behaviour at baseline, and 56% (n = 23) and 34% (n = 14) provided validated accelerometer data at follow-up 1 and follow-up 2 time points, respectively. Across all time points, 31% (n = 20) of accelerometer data sets were assessed as invalid. The main reason for invalidity was that the device was not worn (35%).

• Information about diagnosis (e.g. tumour size, lymph nodes and metastases) was not available for all participants, thereby making it difficult to compare diagnosis between intervention and control groups.

• There was recruitment bias; at baseline, participants were meeting or nearly meeting recommended levels for physical activity (i.e. 30 minutes of moderate physical activity per day), had good self-reported quality of life and low levels of fatigue and low anxiety and depression, high physical activity self-efficacy and risk perception.

• Analyses suggest no intervention effect on outcomes, but given that the study was not designed to evaluate health outcomes, had a small sample size and poor completion rate for the primary outcome (i.e. physical activity), this finding is not reliable.

• Intervention fidelity was compromised because no additional cancer-related education sessions were provided for CRC participants attending cardiac rehabilitation.

Strengths and limitations of phase 2

People with CRC who agreed to participate in this study may be particularly keen to increase their level of physical activity, which means that the findings from CRIB may not be applicable to people with CRC who are likely to be less interested in being physically active to aid their recovery and reduce risk of recurrence. In addition, this pilot trial was small scale, recruiting patients with CRC from only three UK hospitals. Nonetheless, this is the first time that an already existing rehabilitation service (i.e. cardiac rehabilitation) has been pilot tested for people with cancer. In this respect, these findings are novel and can be used to inform future research directions. Importantly, phase 2 highlights ways in which trial procedures can be improved for a future large-scale trial to measure effectiveness.

Key trial parameters

In this chapter we discuss the CRIB study in relation to similar studies. Thus, Table 37 compares key trial parameters of physical activity interventions for people with CRC. Although we make comparisons between studies, CRIB was the only pragmatic trial. We tested an existing NHS service (i.e. cardiac rehabilitation) in real-world settings, whereas all of the other trials were experimental, with an intervention specifically designed and controlled by the research team. Studies use different definitions to report rates and so we applied the following definitions and calculations for the purposes of comparison:

• Design of the study was defined as a RCT, a non-randomised trial or a before-and-after study of the intervention.

• Mode of the intervention was defined as exercise classes, exercise counselling (telephone or face to face) or home-based exercise prescription (i.e. participants are given physical activity goals at the start of the intervention and expected to meet these goals without any or with minimal contact with an instructor).

• Registered patients referred to the pool of people with CRC that potentially eligible participants were identified from through screening (e.g. in CRIB this was the number of people with CRC admitted to hospital for surgery).

• Assessed for eligibility was defined as the number of people with CRC who were assessed for eligibility using study inclusion/exclusion criteria.

• Eligibility rate was calculated by dividing the number of people with CRC who were assessed for eligibility by the number who met inclusion criteria (note this is different from how we defined eligibility rate in Chapter 7, which was calculated by dividing the number of people with CRC admitted for surgery by the number who met eligibility criteria). The calculation was changed for Table 37 because so few studies report the number of patients potentially eligible; instead, they report the number screened for eligibility, which is generally a considerably smaller number. Making the change allowed us to compare more studies.

• Consent rate was calculated by dividing the number of people with CRC who met inclusion criteria (i.e. eligible) by the number who consented to participate in the study.

• Completion rate was calculated by dividing the number of participants who had consented to participate in the study by the number of participants who completed outcome measures at different time-points (e.g. baseline and follow-ups).

• Missing data were defined as the number of participants not entered into analyses because of invalid accelerometer data.

• Intervention adherence was calculated by dividing the number of planned physical activity sessions/consultations by the number actually attended.

Sample size calculation for a definitive randomised controlled trial

This study did not set out to measure the effectiveness of the intervention; rather, it was conducted to find out if cardiac rehabilitation is a feasible and acceptable rehabilitation service for people with CRC and to gather information to hone trial procedures for a future effectiveness trial. The feasibility work has provided process information that can inform a future trial but the effect data are insufficient to support a robust sample size calculation for a future definitive RCT. This is due to a small number of eligible participants being recruited (n = 41), decreasing completion rates at follow-up (61% at final follow-up), missing data (31%) and a recruitment bias of physically active and healthy participants.

Recommendations for improving trial procedures for a future effectiveness trial

Phase 2 suggests that it is feasible to conduct a definitive trial. Nonetheless, based on the results of phase 2, Table 42 presents proposed recommended changes to main trial components and their estimated impact.

Introduction

Qualitative methods are an essential part of a trial’s evaluation and particularly apt for exploring the feasibility and acceptability of trial components and the intervention as opposed to measuring outcomes. Qualitative methods in RCTs can be used to understand and improve main trial components, such as recruitment. 164 Qualitative methods can also be used to explore processes, contextual factors or intervention characteristics and mechanisms that can aid the interpretation of trial outcomes. 165 Nevertheless, qualitative studies as an embedded component of RCTs remain uncommon,166 although there has been a recent growth in use of qualitative methods under the auspices of the MRC Collaboration and Innovation for Difficult or Complex Randomised Controlled Trials (ConDuCT) Hub. 167–170

The phase 2 qualitative study nested within the pilot RCT explored participants’ and clinicians’ views and experiences of the main components of the trial and the intervention (i.e. cardiac rehabilitation). The specific aims were to investigate:

• the views and experiences of CRC participants of the main trial procedures (e.g. recruitment, randomisation)

• the views and experiences of cardiac rehabilitation as a feasible and acceptable rehabilitation programme for CRC from the perspectives of CRC participants, people with CHD attending cardiac rehabilitation, and cardiac rehabilitation clinicians and CRC nurse specialists.

We have used the COnsolidated criteria for REporting Qualitative research, a 32-item checklist for interviews and focus groups,171 to guide the structure of this chapter.

Research team

Four investigators collected qualitative data by interview and focus group. All four investigators attended a 1-day training course in conducting qualitative interviews organised by the Social Research Association Scotland and one of the investigators involved in data analysis also attended 1-day training in qualitative data analysis. None of the investigators conducting interviews and focus groups was involved in providing patient care. Three were employed on the study as research assistants employed by the University of Stirling and one was a research nurse employed by the NHS trust.

Study design

Findings

Discussion

Economic evaluation

The main aims of the economic evaluation component in the CRIB pilot study were to identify the main resource implications of delivering the rehabilitation programme and the subsequent impact on NHS care required by the participants; and devise and test approaches and tools to measure these.

To do this, the analysis looked at:

1. response rates to data collection tools

2. the cost of delivery for the rehabilitation groups

3. the health-care costs of participants in both trial arms

4. health-related quality of life through the calculation of utility values using the EQ-5D-5 Levels (EQ-5D-5L).

Feasibility

The main purpose of the CRIB pilot trial was to prepare the way for a full trial by testing methods of delivery and trial materials. In this pilot trial, response rates for the Health Resource Use questionnaire and EQ-5D-5L were high, and it appears that patients did not have problems filling in the questionnaires. A total of 40 (97.5%) (control, n = 20; intervention, n = 20) out of 41 participants, and 31 (75.6%) (control, n = 16; intervention, n = 15) and 25 (61%) (control, n = 13; intervention, n = 12) participants completed the questionnaires at the follow-up 1 and follow-up 2 time points, respectively. Overall, 60.9% of participants completed the trial.

Cost data

The overall cost of the programme is divided into two main components: the cost of delivery of the rehabilitation groups and the health-care resource use of all trial participants.

Discussion of economic evaluation findings

Economic evaluation can be defined as a comparison of alternative options in terms of costs and consequences. 201 As such, one can argue that we do not perform a full economic evaluation here as we mainly present costs and consequences rather than perform a formal comparison between interventions. Given the limited sample size and no difference in the primary clinical outcome measure and in measured health gain, as captured by the EQ-5D, between the two groups, we decided not to perform a full economic analysis in this pilot trial. In a future main trial, a cost-effectiveness analysis will be performed following methods as outlined by Drummond et al. 201 Costs and outcomes for each of the trial participants will be calculated and the mean incremental cost-effectiveness ratio (ICER) will be estimated as:

The primary economic outcome in this analysis will be the quality-adjusted life-year, using the EQ-5D-5L, measured at different time points during the trial, and the economic evaluation will estimate the incremental cost per quality-adjusted life-year associated with the treatment as stated in the protocol and shown above. The additional (incremental) costs associated with the treatment, when added to usual care, will be estimated using resource-use data collected within-trial, and unit costs for resource use from national published/NHS sources.

Hence, we are pleased to report that in this pilot all trial materials and questionnaires were tested successfully and proved suitable for economic evaluation purposes. There was a high response rate for Bristol Online Survey which included EQ-5D and other measures used in the trial. There were no missing data beyond people who did not participate in the follow-ups. We recognise that recall bias is a problem when using self-report to measure health service use. For the future we recommend that the larger full trial will need to clearly establish the ‘true’ costs of health service use using objective measurement and of the rehabilitation group and whether or not these can absorbed into current practice. It was initially planned that the CRC participants would enter existing classes, keeping costs at a minimum, so we need to establish if this is possible. A full economic evaluation during the main trial will require accurate cost data, so this issue must be addressed. Nevertheless, based on the data reported in this section and elsewhere in this report, we believe that a full trial is feasible and a full economic evaluation can be performed if similar materials and methods are employed in a future definite trial.

Involvement of Bowel Cancer UK (Scotland)

The manager of Bowel Cancer UK (Scotland), Emma Anderson, was a member of our project advisory group. Two patient advisors who were recruited from Bowel Cancer UK (Scotland) also participated in the study. All three contributed towards the design of the study, advising on, for example, study information sheets and consent forms. One of the patient advisors moved abroad and so curtailed her involvement in the study. The other patient advisor, Gillian Sweetman, presented at the NIHR ‘welcome’ event, was a member of the project advisory group and conducted interviews with three people with CRC during phase 1. She participated in a 1-day training course so that she could conduct these interviews. Dr Gill Hubbard (the study’s principal investigator), who organised the event for members of the Service User Research Partnership of Breast Cancer Care, delivered the training. Gillian Sweetman was invited with agreement from Breast Cancer Care.

In March 2015, Gillian recounted her overall experience of being involved in the CRIB project (Box 5).

Discussion and recommendations

In this study, we involved a leading UK CRC charity. We worked particularly closely with a full-time member of staff and a volunteer who had direct and personal experience of CRC as a patient. Their views about the study, in terms of its importance for people with CRC, were articulated at research group meetings, which instilled confidence that the study was worthwhile and would be supported by the charity at the national level. In addition, Gillian’s involvement demonstrated that, with training and support, volunteers can work closely with the researchers, helping, for instance, to collect data. This helped to situate Gillian as one of the team rather than as the patient representative on the team. The difference between the two approaches may be subtle, but by involving volunteers as co-researchers in this way, we developed a more inclusive team. Based on our experience, this pilot work can be used as a springboard for further PPI in any future trial.

Results

Limitations

This study was conducted to inform design and make decisions for a future full-scale trial of cardiac rehabilitation for cancer patients, including whether or not a full-scale trial should be done at all. This requires careful judgements about the limitations of this study and the impact those limitations will have on the applicability of what we have found to any future full-scale trial. These judgements are part of all feasibility and pilot studies. Broadly speaking, potential limitations fall into three categories:

1. Design and conduct challenges: challenges and problems encountered in the feasibility and pilot study that suggest changes are needed if a full-scale trial is to be feasible.

2. Unanticipated conduct failures: a planned part of the feasibility and pilot study was not possible because of some external event not linked to the intervention under evaluation (e.g. slow approvals, loss of trial staff).

3. Generalisability of the sites selected for the feasibility and pilot work: whether or not there is reason to suggest that the sites involved in the feasibility and pilot study are atypical of sites that would be part of a full-scale trial.

Item 1 is what feasibility and pilot studies are chiefly designed to identify: issues that make a full-scale trial either unfeasible or likely to be unfeasible without modification to what was planned and subsequently used in the feasibility study. In our study, there were many trial design issues falling into this category, including suboptimal eligibility, consent and completion rates and missing data. Issues relating to the intervention – cardiac rehabilitation – falling into this category include barriers to attending cardiac rehabilitation and the capability of existing cardiac rehabilitation clinicians to support people with cancer.

To address these challenges and study limitations we recommend induction training for staff involved in recruitment and data collection. The training, for instance, would address suboptimal recruitment by highlighting to staff the health benefits of cardiac rehabilitation for patients who have poor mobility with the aim of boosting the recruitment of those patients in relatively poorer health. The training would also aim to address missing data by increasing researchers’ competence to collect accelerometer data by for instance, emphasising the importance of these data for the study and showing staff how to explain to participants how and when to wear the accelerometer device. To improve the consent rate we propose to remove a major barrier to participation in the study, travel to and distance from cardiac rehabilitation, by including outreach services in any future trial. Offering cardiac rehabilitation outreach services would also address a limitation of the intervention, which was that cardiac rehabilitation was too far away for some people to travel. Another barrier to consenting to the study and also to attending cardiac rehabilitation highlighted by our feasibility and pilot work was poor recovery from surgery and ongoing treatments, especially adjuvant therapy. Indeed, there was a general reluctance by participants to start exercising during ongoing treatment. To improve the consent rate and address barriers to participating in the cardiac rehabilitation programme, we therefore suggest including only patients who are at the end of treatment. We also recommend adopting a strategy whereby patients are involved in the decision about their start date for attending cardiac rehabilitation. This is because a limitation of the study design, and also of the intervention, was that some patients needed more time post surgery to agree to participate than we provided. Additionally, questionnaire completion rates fell from baseline to first and then second follow up. To improve completion rates we recommend the adoption of evidence-based strategies such as monetary incentives to return questionnaires together with regular reminders. Finally, our study found that cardiac rehabilitation clinicians did not feel competent providing specific cancer-related psychosocial support to participants with CRC. Hence to address this limitation we propose that cancer clinicians should continue to provide cancer-specific psychosocial support while cardiac rehabilitation clinicians provide generic as opposed to disease-specific psychosocial support as well as supporting people to exercise.

Item 2 in the list of limitations differs from item 1, in that the limitation is not necessarily a result of the trial design, the intervention or its delivery, but of the inability of the study team to make something happen because of one or more unanticipated problems: generally an external event, such as the loss of a key member of staff. Whether or not these problems could have been anticipated is a moot point; the key issue is that they were not anticipated, the impact was substantial and, as a consequence, part of the study could not be carried out. In our pilot study, we were unable to meet one study objective, which was to provide data for sample size calculations for a definitive RCT. Our inability to provide data was a consequence of failing to meet our recruitment target of 66 (the recruitment challenges mentioned above meant that we recruited 41) and also of recruitment bias. Indeed, a key finding from the feasibility and pilot work, and also a good illustration of why feasibility and pilot work is important, was recruitment bias; most participants were already meeting or close to meeting the recommended level of physical activity for the adult population (i.e. 150 minutes per week of moderate physical activity), which is unusual for this patient group and, indeed, for the general population. 14–19 To address this study limitation we propose in a future trial to revise the eligibility criteria and exclude patients who are already physically active. This would also mean that the cardiac rehabilitation would be targeting those patients who need it most, that is, those who are currently not meeting the recommended level for physical activity associated with health benefits. This may, of course, extend the recruitment period of any full-scale trial, but this could be planned for from the start.

Item 3 is a challenge for all feasibility and pilot studies, and perhaps triallists have a tendency to overestimate the generalisability of findings from small-scale work carried out in a handful of sites to what might be many tens of sites in a future full-scale trial. There is, of course, a balance to be struck between the duration and cost of feasibility and pilot work and the duration and cost of a potential full-scale trial. For example, recruitment is still an important concern for all triallists; a recent study of NIHR- and MRC-funded trials found that about half of these trials met their targets despite some of the trial teams having done feasibility and pilot work. 204 An earlier study that looked explicitly at the impact of pilot work on recruitment did not find a clear link between doing pilot work and successful recruitment. 205 Our feasibility and pilot work suggests variation in sites regarding recruitment parameters (we have written about this in detail elsewhere141 as well as in this report) and we therefore anticipate that some sites will meet recruitment targets and others will not. Which sites are likely to perform well is difficult to judge in advance. It is for this reason, in particular, that we propose an internal pilot as part of any future multisite effectiveness trial. The internal pilot should have clearly defined stop–proceed rules based around recruitment targets. Another limitation of the study is that people with CRC who agreed to participate might have been particularly keen to increase their level of physical activity, which means that the findings may not be applicable to those people with CRC likely to be less interested in being physically active to aid their recovery and reduce risk of recurrence. As discussed above, we are proposing that any future trial should exclude those who are already physically active. Limitations on the ability to make generalisations about the intervention are that the costs associated with the referral of cancer patients to cardiac rehabilitation is likely to vary by site; the capacity of a cardiac rehabilitation service to take additional patients is also likely to vary from one site to the next. These variations are likely to be important information for those who are tasked with commissioning or managing services. Finally, we recommend embedding a process evaluation in any future trial so that contextual factors impacting the delivery of cardiac rehabilitation to patients with cancer (including the capacity to take more patient referrals) in each site are also highlighted. This is important given that there are over 300 cardiac rehabilitation services in the UK, and this and any future study is likely to include only a fraction of these sites.

In summary, this feasibility and pilot work highlights a range of trial design limitations, including suboptimal eligibility, consent and completion rates, missing data and recruitment bias. It also highlights the limitations of cardiac rehabilitation for patients with cancer, including capacity, costs and capability issues. To make a full trial feasible, we have made a series of recommendations to address the limitations we have identified, including an internal pilot with clear stop–proceed rules, induction training for staff and participant incentives. To clarify and aid interpretation of generalisation, we also recommend an embedded process evaluation so that each site’s contextual factors impacting cardiac rehabilitation for patients with cancer are illuminated.

Conclusions

Contributions of authors

Dr Gill Hubbard is an experienced research project manager in cancer care and rehabilitation. Dr Hubbard was chief investigator and was involved in all aspects of the study, from project conception through to report completion and submission.

Ms Julie Munro is a researcher in exercise rehabilitation and was involved in trials management of the three sites throughout data collection. Ms Munro was also involved in drafting and editing all sections of the report.

Professor Ronan O’Carroll is a behavioural scientist who was involved in protocol development and contributed to edits of the final report.

Professor Nanette Mutrie (MBE) is a physical activity expert was involved in protocol development and contributed to edits of the final report.

Dr Lisa Kidd is a specialist in self-care in cancer and qualitative research, was involved in protocol development and contributed to edits of the final report.

Professor Sally Haw has expertise in public health, research methods and the evaluation of complex interventions, and was involved in protocol development and contributed to edits of the final report.

Dr Richard Adams is a senior lecturer and consultant in lower gastrointestinal oncology, with extensive experience in phase II and III colorectal clinical trials. Dr Adams was involved in protocol development, facilitating study set-up on one site with support and contributed to edits of the final report.

Professor Angus JM Watson is consultant in colorectal and general surgery with extensive research input and experience. Professor Watson commented on edits of the final report.

Professor Stephen J Leslie is consultant cardiologist with an extensive research portfolio. Professor Leslie was involved in protocol development and contributed to edits of the final reports.

Ms Petra Rauchhaus is an experienced clinical trials statistician who was involved in protocol development. Ms Rauchhaus produced the statistical analysis plan and completed data analysis for all report outcomes.

Dr Anna Campbell (MBE) is an expert in physical activity, with extensive experience in the field of cancer and exercise. Dr Campbell was involved in protocol development and contributed to edits of the final reports.

Dr Helen Mason is a senior lecturer in health economics. Dr Mason produced Chapter 10, the economic evaluation, for the report.

Dr Sarkis Manoukian is a researcher in health economics, involved in a number of clinical trials. Dr Manoukain produced Chapter 10, the economic evaluation, for the report.

Dr Gillian Sweetman was our patient advisor, who went on to contribute at steering group meetings. Dr Sweetman collected qualitative data and drafted and contributed to Chapter 11 of the report.

Professor Shaun Treweek is an experienced health services researcher with expertise in trial design, the development of complex interventions and trial recruitment. Professor Treweek was involved in protocol development and contributed to edits of the final report.

Publications

Munro J, Adams R, Campbell A, Campbell S, Donaldson C, Godwin J, et al. CRIB – the use of cardiac rehabilitation services to aid the recovery of patients with bowel cancer: a pilot randomised controlled trial (RCT) with embedded feasibility study. BMJ Open 2014;4:e004684.

Hubbard G, Campbell A, Davies Z, Munro J, Ireland AV, Leslie S, et al. Experiences of recruiting to a pilot trial of cardiac rehabilitation in patients with bowel cancer (CRIB) with an embedded process evaluation: lessons learned to improve recruitment. Pilot Feasibility Stud 2015;1:15.

Hubbard G, Adams R, Campbell A, Kidd L, Leslie S, Munro J. Is referral of postsurgical colorectal cancer survivors to cardiac rehabilitation feasible and acceptable? A pragmatic pilot randomised controlled trial with embedded qualitative study. BMJ Open 2016;6:e009284.

Hubbard G, O’Carroll R, Munro J, Mutrie N, Haw S, Mason H, Treweek S. The feasibility and acceptability of trial procedures for a pragmatic randomised controlled trial of a structured physical activity intervention for people diagnosed with colorectal cancer: findings from a pilot trial of cardiac rehabilitation vs. usual care (no rehabilitation) with an embedded qualitative study. BMC Pilot Feasibility Stud 2016; in press.

Data sharing statement

All available data can be obtained from the corresponding author: Gill Hubbard, Cancer Care Research Centre, School of Health Sciences, University of Stirling, Highland Campus, Centre for Health Science, Old Perth Road, Inverness, IV2 3JH. Telephone: + 44 (0) 1463 255649. E-mail: gill.hubbard@stir.ac.uk.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HS&DR programme or the Department of Health. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HS&DR programme or the Department of Health.