For Peer Review Only

BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

An interactive web-based Pulmonary Rehabilitation programme: A randomised controlled feasibility trial.

ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2016-013682

Article Type: Research

Date Submitted by the Author: 16-Aug-2016

Complete List of Authors: Chaplin, Emma; University Hospitals of Leicester NHS Trust, Pulmonary Rehabilitation Hewitt, Stacey; Centre for Exercise and ehabilitation Science, Leicester Respiratory Biomedical Research Unit Apps, Lindsay; Centre for Exercise and ehabilitation Science, Leicester Respiratory Biomedical Research Unit Bankart, M. John; Keele University, Institute of Primary Care Pulikottil-Jacob, Ruth; Warwick Medical School , Warwick Evidence Boyce, Sally; Centre for Exercise and ehabilitation Science, Leicester Respiratory Biomedical Research Unit Morgan, Mike; University Hospitals of Leicester NHS Trust, Respiratory Medicine Williams, Johanna; Glenfield Hospital, University Hospitals of Leicester NHS Trust, Respirartory Medicine Singh, Sally; University Hospitals of Leicester NHS Trust, http://bmjopen.bmj.com/ Cardiac/Pulmonary Rehabilitation

Primary Subject Rehabilitation medicine Heading:

Secondary Subject Heading: Respiratory medicine

SPACE for COPD, chronic obstructive pulmonary disease, Internet, Web- Keywords:

based, pulmonary rehabilitation on September 23, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 An interactive web-based Pulmonary Rehabilitation programme: A randomised controlled 4 feasibility trial. 5 6 Chaplin, Emma1, Hewitt, Stacey1, Apps, Lindsay1, Bankart, John2, Pulikottil-Jacob, Ruth3, 7 1 1 1 1,4 8 Boyce, Sally , Morgan, Mike , Williams, Johanna , Singh, Sally 1 9 Centre for Exercise and Rehabilitation Science, Leicester Respiratory Biomedical Research Unit, Department of 10 Respiratory Medicine, Glenfield Hospital, University Hospitals of Leicester, Leicester LE3 9QP UK 2 11 Department of Primary Care and Health Sciences, Keele University, Keele, UK 3 12 Health Sciences Research Institute, Medical School, University of Warwick, Coventry, UK 4 13 School of Sport, Exercise and Health Sciences, Loughborough University, Loughborough, UK 14 15 For peer review only 16 17 Corresponding Author: Emma Chaplin, BSc (Hons), Centre for Exercise and Rehabilitation 18 Science, NIHR Leicester Respiratory BRU, University Hospitals of Leicester NHS Trust, 19 Leicester, LE3 9QP, UK. +44116 2583181, [email protected] 20 21 Word Count: 22

23 24 Key words: SPACE for COPD; Internet; web-based; chronic obstructive pulmonary disease; 25 pulmonary rehabilitation 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 23, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 28 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 Abstract: 4 5 Objectives: The aim of this study was to determine if an interactive web-based Pulmonary 6 7 Rehabilitation (PR) programme is a feasible alternative to conventional PR. 8 9 Design: Randomised controlled feasibility trial 10 11 Setting: Participants were recruited from PR assessments, primary care and community 12 rehabilitation programmes. Patients randomised to conventional rehabilitation commenced 13 14 the programme according to the standard care at their referred site on the next available 15 date. For peer review only 16 17 Participants: 103 patients were recruited to the study and randomised: 52 to conventional 18 rehabilitation [mean (±SD) age 66 (±8) years, MRC 3 (IQR2-4)]; 51 to the web arm [mean 19 20 (±SD) age 66 (±10) years, MRC 3 (IQR2-4)]. Participants had to be willing to participate in 21 either arm of the trial, have internet access and be web literate. 22 23 Interventions: Patients randomised to the web-based programme worked through the 24 website, exercising and recording their progress as well as reading the educational material 25 provided, on line. Conventional PR consisted of twice weekly, two hourly sessions (an hour 26 27 for exercise training and an hour for an education). 28 29 Outcome measures: Recruitment rates, eligibility and patient preference as well as data on 30 drop out and completion rates for both programmes were collected. Standard outcomes for 31 a PR assessment including measures of exercise capacity and quality of life questionnaires 32 33 were also evaluated. http://bmjopen.bmj.com/ 34 35 Results: A statistically significant improvement (p≤ 0.01) was observed within each group in 36 the ESWT (WEB: mean change 189 ± 211.1; PR classes: mean change 184.5 ± 247.4secs) and 37 CRQ-D (WEB: mean change 0.7 ± 1.2; PR classes: mean change 0.8 ± 1.0). However there 38 were no significant differences between the groups in any outcome. 39 40 41 Conclusion: An interactive web-based PR programme is feasible and acceptable when 42 compared to conventional PR. Future trials maybe around choice based PR programmes for on September 23, 2021 by guest. Protected copyright. 43 select patients enabling stratification of patient care. 44 45 Trial registration number ISRCTN03142263 46 47 48 49 Strengths and Limitations of this study 50 • The study concentrates on the feasibility of an interactive web-based Pulmonary 51 Rehabilitation programme (SPACE: Self-management Program of Activity, Coping 52 53 and Education for COPD®). It will provide data on recruitment, eligibility and 54 patient preference which will inform future trials around choice based 55 programmes for select patients. 56 • Much of the success depends on the eligibility of the patients, i.e. web literacy and 57 access to the internet. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 • The study will compare a variety of clinical outcomes between a web-based and a 4 conventional rehabilitation PR programme in order to test the various components 5 of the intervention. 6 7 8 9 10 Introduction 11 12 Chronic Obstructive Pulmonary Disease is the fourth leading cause of death in the United 13 14 15 Kingdom (UK)For and is characterised peer by a progressivereview deterioration only of debilitating symptoms 16 17 and increasingly frequent exacerbations. Pulmonary rehabilitation (PR) has been proven to 18 19 20 be effective in improving quality of life, psychological functioning and physical activity and 21 22 national guidelines recommend that PR should be offered and made available to all those 23 24 with COPD (1). The standard provision of PR is a supervised package of exercise and 25 26 27 education usually twice a week for a minimum of six weeks, which is either hospital or 28 29 community based, and supported by a home exercise programme (2). However the barriers 30 31 to uptake of a PR programme have previously been reported (3) which included transport, 32 33 http://bmjopen.bmj.com/ 34 the perceived benefits of PR, disruption to usual routine and the timings of programmes. 35 36 These factors play some contribution as to why programmes have poor attendance and 37 38 adherence resulting in many of the programmes reporting dropout rates as high as 50%. 39 40 41 42 Ongoing changes and challenges means that the NHS and the services it provides need to on September 23, 2021 by guest. Protected copyright. 43 44 adapt to take advantage of and capitalise on the opportunities that new technologies and 45 46 47 treatments can offer to patients (4). There is a growing evidence base for the use of the 48 49 internet in the management of many chronic conditions in areas as diverse as the 50 51 management of diabetes, Parkinson’s disease, depression, rheumatoid arthritis, asthma, 52 53 54 chronic pain and epilepsy (5-11). Computer-tailored interventions have been shown to 55 56 effectively improve health behaviours such as physical activity (12) and be cost effective 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 28 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 (13). The provision of a menu based and patient centred service is said to be essential to 4 5 improve uptake and completion rates within Cardiac Rehabilitation (CR) (14). However at 6 7 8 present there is no choice within Pulmonary Rehabilitation. A web-based PR programme has 9 10 the potential to be a novel and effective approach to increasing patient choice in the mode 11 12 of delivery and setting of rehabilitation (especially to those patients who decline the offer of 13 14 15 conventionalFor pulmonary peerrehabilitation) whilereview simultaneously increasingonly the capacity of PR. 16 17 18 We have previously developed and described ‘ACTIVATE YOUR HEART®’ (AYH)(15; 16) which 19 20 21 is an interactive web-based cardiac rehabilitation (CR) programme that has proved very 22 23 popular with patients. Brough et al (2014) reported a significant improvement in exercise 24 25 capacity and quality of life in patients that completed the web-based programme. Following 26 27 28 the success of AYH, we have developed a prototype website based on the educational 29 30 content of the ‘SPACE for COPD®’ self-management workbook. SPACE (Self-management 31 32 programme of Activity, Coping and Education) for COPD® (17) is a structured programme of 33 http://bmjopen.bmj.com/ 34 35 exercise, education and psychosocial support which has been developed by our institution 36 37 as a collaboration between experts, patients and carers and has been awarded a Crystal 38 39 40 Mark for Clarity by the Plain English Campaign (18). 41 42 on September 23, 2021 by guest. Protected copyright. 43 The aim of this feasibility study was to provide quantitative, economic and technical data to 44 45 see if an interactive web-based PR programme was a feasible alternative compared with 46 47 48 conventional PR. This included: 49 50 1. Gathering information regarding the recruitment rate of patients who were eligible and 51 52 willing to be randomised to either the web-based or the conventional rehabilitation 53 54 55 programme, and to monitor retention and drop out through all stages of the programmes. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 2. Comparing a variety of clinical outcomes between a web-based and a conventional 4 5 rehabilitation PR programme in order to test out the various components of the 6 7 8 intervention and identify any technical or other difficulties that may be inherent in the 9 10 delivery of a web-based PR programme. 11 12 13 14 Methods 15 For peer review only 16 Participants 17 18 Eligibility criteria for participants 19 20 21 Eligible participants had an established diagnosis of Chronic Obstructive Pulmonary Disease 22 23 (COPD) defined as a forced expiratory volume in one second (FEV1),post-bronchodilation of 24 25 <80% and a predicted ratio of FEV1 to forced vital capacity of 0.70 and a Medical Research 26 27 28 Council (MRC) dyspnoea score (19) of between 2 and 5. Patients had to be willing to partake 29 30 in either arm of the study. Access to the internet for more than 3 months, the ability to 31 32 navigate around a variety of websites (for example, uses online shopping or banking 33 http://bmjopen.bmj.com/ 34 35 websites) and regular use of email was required. Patients also had to be able to read and 36 37 write in English. 38 39 40 41 42 Patients were excluded if they were unable to participate in the exercise component of the on September 23, 2021 by guest. Protected copyright. 43 44 rehabilitation programme due to other comorbidities or had done PR in the previous 12 45 46 47 months. Eligible patients had to be willing and able to take part in the web-based 48 49 programme. 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 28 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 Setting 4 5 6 Participants to the study were primarily recruited from those patients that had been 7 8 9 referred for pulmonary rehabilitation at University Hospitals of Leicester (UHL) NHS Trust. 10 11 Recruitment was also directly from primary care and community rehabilitation services 12 13 within Leicester Partnership Trust (LPT) and eligible participants were identified from the 14 15 For peer review only 16 research participant database of the Leicester Respiratory Biomedical Research Unit and 17 18 Pulmonary Rehabilitation Department. 19 20 21 22 23 Randomisation 24 25 Patients were randomised to either the conventional rehabilitation programme as is 26 27 28 standard at their referred site or the web-based Pulmonary Rehabilitation programme 29 30 (SPACE for COPD®). Randomisation to the treatment group allocation was on a 1:1 ratio to 31 32 either group and was performed using a web based programme 33 http://bmjopen.bmj.com/ 34 35 (www.sealedenvelope.com). 36 37 38 39 Trial Interventions 40 41 42 Intervention group – web-based pulmonary rehabilitation programme on September 23, 2021 by guest. Protected copyright. 43 44 Following randomisation to the intervention group, patients attended a standardised 45 46 introductory session where participants were given a password protected secure log-in to 47 48 49 the website as well as written instructions on website navigation. 50 51 52 53 54 Patients were signposted to all the relevant sections on the website including the home 55 56 exercise programme and goal setting. There was also an individualised webpage (Figure 1) 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 featuring a personalised action plan designed to assist in the management of exacerbations 4 5 which was completed by the rehabilitation specialist in conjunction with the patient. 6 7 8 9 10 As in conventional pulmonary rehabilitation, patients were encouraged to exercise on a 11 12 daily basis at home and record their progress in the online exercise diary section. 13 14 15 Throughout Forthe duration peer of the web-based review programme the patient’s only progress was reviewed 16 17 online and there was weekly contact between the patient and the rehabilitation specialist 18 19 20 via email or telephone using a standardised proforma. This ensured that patients were 21 22 helped to progress their exercise programme appropriately and to answer any queries that 23 24 arose. 25 26 27 28 29 The educational content of the web-based programme was based on the ‘SPACE for COPD®’ 30 31 manual. Patients worked through the website content at their own pace, however certain 32 33 http://bmjopen.bmj.com/ 34 milestones needed to be completed or achieved before further content could be accessed in 35 36 order to ensure appropriate progress through the programme (see supplement for 37 38 description of the different stages on the website). It was anticipated from previous work 39 40 41 (14) that it would take approximately 6-8 weeks to work through the online programme. 42 on September 23, 2021 by guest. Protected copyright. 43 44 45 46 Standard care group - conventional pulmonary rehabilitation programme 47 48 Patients randomised to standard care commenced conventional rehabilitation according to 49 50 the standard care at their referred site which was either hospital or community based. 51 52 53 Conventional pulmonary rehabilitation programmes consisted of twice weekly sessions each 54 55 lasting two hours which were divided into an hour for exercise training and an hour for an 56 57 education session covering a variety of relevant self-management topics. The exercise 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 28 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 training consisted of endurance based walking, static cycling and strength training with 4 5 weights. Patients were progressed through the exercise programme as able and 6 7 8 appropriate. Patients were encouraged to also complete a home exercise programme on 9 10 the days when they did not attend rehabilitation classes and to fill in an exercise diary. The 11 12 educational sessions were conducted as group sessions and delivered by experts in their 13 14 15 field. Topics Forincluded medication, peer relaxation review skills, chest clearance only and breathlessness 16 17 management and energy conservation. 18 19 20 21 22 Outcome measures 23 24 25 All the measures used and collected in the trial including clinical (the incremental (ISWT)(20) 26 27 28 and endurance (ESWT)(21) shuttle walk tests; chronic respiratory questionnaire- self 29 30 reported (CRQ-SR)(22), Hospital Anxiety and Depression Scale (HADS)(23), COPD 31 32 33 Assessment Tool (CAT)(24), PR Adapted Index of Self-Efficacy (PRAISE)(25), Bristol COPD http://bmjopen.bmj.com/ 34 35 Knowledge questionnaire (BCKQ)(26), Euro-QOL (EQ-5D-5L)(27), patient cost questionnaire 36 37 (28)) and non-clinical have previously been described in the study protocol (29). Clinical 38 39 40 measures were performed at baseline and repeated again at the discharge assessment 41 42 following completion of either rehabilitation programme (usually approximately 6-7 weeks on September 23, 2021 by guest. Protected copyright. 43 44 after starting the programme) and were conducted by a research physiotherapist who was 45 46 47 blinded to treatment group allocation. Those patients randomised to the website were 48 49 offered conventional PR classes if they felt it would be more beneficial at discharge. 50 51 Non-clinical outcomes included a web-usage audit for the internet-based programme, 52 53 54 recruitment rates, eligibility and patient preference as well as drop out and completion 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 rates in both treatment groups. Any serious adverse events were reported to the sponsor 4 5 and the patients’ ability to exercise safely was monitored. 6 7 8 9 10 Quantitative Data Analysis 11 12 Data was entered and stored on a secure web-based system (REDCAP) which has 13 14 15 discrepancy Formanagement peer features. Data review was then transferred onlyfrom REDCAP to the 16 17 Statistical Package for the Social Sciences (SPSS) version 18 (SPSS Ltd, Woking, Surrey, UK). 18 19 20 The data were checked for normality before baseline characteristics were compared 21 22 between-groups using an independent t-test. Analysis was primarily descriptive i.e. 23 24 estimation of means and standard deviations, proportion of patients eligible/ willing to 25 26 27 participate in the study. A paired t-test was used to compare within-group changes and an 28 29 independent t-test was used to compare the differences between the two treatment groups 30 31 in the ISWT, ESWT and CRQ-D at the two different time points. 32 33 http://bmjopen.bmj.com/ 34 Health economic evaluation 35 36 The economic analysis was conducted from the perspective of the NHS and personal social 37 38 service and aimed to collect data on resource use within the healthcare system and quality 39 40 41 of life associated with the interventions. The validity and feasibility of administering the 42 on September 23, 2021 by guest. Protected copyright. 43 ‘Patient Cost Questionnaire’ as well as the EQ-5D-5L to inform any future economic 44 45 46 evaluation was ascertained. 47 48 49 Results 50 51 103 patients were recruited and randomised to the study between May 2013 and July 2015: 52 53 54 52 to the conventional PR group and 51 to the web group. Figure 2 shows the flow of 55 56 eligibility, screening, randomisation and follow-up in the study. No significant differences 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 28 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 between the groups’ baseline characteristics or outcome measures (Table 1) were seen. 4 5 More patients dropped out from the web intervention group (n = 29) but there were no 6 7 8 significant differences between the baseline characteristics of those patients that completed 9 10 the programme and those that withdrew. Reasons for dropouts are listed in Figure 2. 11 12 Clinical Outcome measures 13 14 15 Clinical outcomesFor are reported peer in table 2.review A statistically significant only improvement (p≤0.01) 16 17 was observed within each group in the ESWT (WEB: mean change 189 ± 211.1; PR classes: 18 19 20 mean change 184.5 ± 247.4 secs) (Figure 3) and CRQ-D (WEB: mean change 0.7 ± 1.2; PR 21 22 classes: mean change 0.8 ± 1.0) (Figure 4). There were no significant differences between 23 24 the groups in any clinical outcome. All outcome measures used were feasible to administer. 25 26 27 Non Clinical study outcomes 28 29 The average number of weeks to complete the website was 11 ± 4. The stage at which 30 31 patients dropped out at are listed in Table 3. The majority of patients dropped out at the 32 33 http://bmjopen.bmj.com/ 34 beginning of the web programme which suggests once the patient was engaged with the 35 36 programme, they were able to complete it. Those that dropped out tended to be mostly 37 38 MRC 3, had a lower baseline ISWT and a significantly higher HADS anxiety score at baseline 39 40 41 compared with those that completed the web programme. In those patients who had been 42 on September 23, 2021 by guest. Protected copyright. 43 randomised to the PR classes, the main reasons that they were withdrawn, were the 44 45 46 patients not starting the classes and being unable to contact the patient (n=6). 47 48 When patients were asked their treatment preference prior to being randomised, the 49 50 largest proportion of patients wanted the web programme (n = 38%)( Figure 5a & b). Of the 51 52 53 22 patients that completed the web programme, only 3 patients (n=14%) felt they would 54 55 like to attend the PR classes. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 Health Economics 4 5 The generic health related quality of life (HRQoL) was measured using EQ-5D-5L. The mean 6 7 8 EQ-5D utility scores for both groups at baseline and discharge are presented in Table 4.The 9 10 mean EQ-5D score was slightly higher in the intervention than in the standard care group at 11 12 discharge. These scores however did not equate to the intervention being more beneficial 13 14 15 to the patients’For HRQoL. peer review only 16 17 Complete data on EQ-5D were available for 57 participants at baseline (37 standard care, 20 18 19 20 intervention), and 56 participants for follow-up visit (37 standard care, 19 intervention); 21 22 complete data on health service utilisation were available for 40 (14 web based and 26 usual 23 24 care) patients at baseline and 42 (16 web based and 26 usual care) patients at discharge. 25 26 27 Overall the current study showed that it was feasible to collect resource utilisation and EQ- 28 29 5D data for a full cost-effectiveness analysis in subsequent trials. 30 31 Discussion 32 33 http://bmjopen.bmj.com/ 34 Pulmonary Rehabilitation is one of the most effective non pharmaceutical therapies for 35 36 patients with COPD, which offers long term benefits. Issues such as transport and locality 37 38 still persist for many patients some of which could be addressed by improving accessibility. 39 40 41 Data from a recent national COPD audit (30) suggests that the capacity of PR programmes is 42 on September 23, 2021 by guest. Protected copyright. 43 inadequate to meet the demand or need. Alternative or more flexible provisions of PR 44 45 46 programmes therefore need to be considered. 47 48 Voncken-brewster et al (2015)(31) tested the effectiveness of a web-based computer- 49 50 tailored COPD self-management intervention on physical activity and smoking behaviour. 51 52 53 There were no statistically significant effects on health related behavioural or clinical 54 55 outcomes. They state this may have been attributed to the low exposure to the application 56 57 and engagement with the program has been shown to be crucial for the effectiveness of 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 28 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 computer tailored interventions. Another randomised controlled trial by Pinnock et al in 4 5 2013 (32) found telemonitoring to not reduce hospital admission or improve patients’ 6 7 8 quality of life. This feasibility study aimed to evaluate the application of an exercise 9 10 intervention as well as promoting self-management. 11 12 The data from this study suggests that an interactive WEB-based programme has the 13 14 15 potential to Forbe a feasible peer and acceptable review alternative when compared only with conventional PR. 16 17 Although both groups improved in the ISWT, it wasn’t significant and did not meet the 18 19 20 minimally clinical important difference (MCID) of 48m (33). The baseline ISWT scores were 21 22 higher in these patients compared with those patients seen normally in our clinical service 23 24 and therefore may account for the small change in the ISWT score. Similar improvements in 25 26 27 time (3 minutes) were seen in the endurance shuttle walk tests for both groups. Health 28 29 related quality of life, measured by the CRQ-SR, appeared to improve significantly, both 30 31 groups exceeding the MCID for the CRQ-D of 0.5. High dropout rates in the web arm of the 32 33 http://bmjopen.bmj.com/ 34 study may have influenced the outcome measures of exercise capacity. The study would 35 36 suggest, like other rehabilitation studies that using a HRQoL measure is a feasible primary 37 38 outcome measure. However assessing for non-inferiority using the CRQ meant a sample size 39 40 41 calculation of 89 patients in each group (accounting for attrition) would have been required. 42 on September 23, 2021 by guest. Protected copyright. 43 Adherence to Web-based programs can vary due to many reasons ranging from lack of time, 44 45 46 to refusing to complete the program. Several features have been identified that could help 47 48 to improve adherence to a web program: making the program tailored to the user and 49 50 interactive (34) as well as allowing users to set personal goals (35). 51 52 53 Initial withdrawals in this study appeared to have been at the exercise stages of the web 54 55 programme as there was a higher dropout at stage 2. This component of the WEB 56 57 programme was simplified based on participant feedback and completion rates improved. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 Similar numbers withdrew after this modification, however these were due to 4 5 exacerbations, other problems from co-morbidities and technology problems meaning 6 7 8 patients didn’t even register or come for their introduction. Although more patients 9 10 withdrew from the web arm of the study (12 from the PR group: 29 from the web group), 11 12 there were no significant differences between the groups in any outcome. Those that did 13 14 15 complete theFor website, didpeer as well as those review that had completed only conventional PR. 16 17 Priorities around the use of technology within the NHS are changing (4) but this will not be 18 19 20 without its challenges. A previous evaluation to explore the use of technology within a 21 22 COPD population done by our institution found although patients owned a computer or 23 24 mobile phone, usage was limited and was predominantly within the younger age range (36). 25 26 27 17% of interested participants in this study did not have access to the internet and 28 29 highlighted the need to assess access as well as the competency of patients being able to 30 31 use the web prior to starting the web programme. The majority of patients that entered the 32 33 http://bmjopen.bmj.com/ 34 study wanted the website showing that there is a desire for this type of intervention. 35 36 However, the study showed a lack of engagement in technology in this particular population 37 38 despite a great deal of patient user involvement in the site development. 39 40 41 The trial design meant that patients needed to be willing to be randomised to either group, 42 on September 23, 2021 by guest. Protected copyright. 43 whereas in clinical practice it is more likely that patients will have a preference due to 44 45 46 genuine choice or practical difficulties that precludes access to supervised rehabilitation 47 48 programmes. This patient choice or preference may improve uptake and completion of a PR 49 50 programme. By exploring alternative forms, such as a web based programme, may mean 51 52 53 those patients that potentially would decline standard PR are provided with a form of 54 55 intervention. Alternative formats of CR, including home-based CR, such as the Heart manual 56 57 and the Angina plan (37), have been shown to be an effective alternative to conventional 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 28 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 CR. In both randomisation groups, when patients were asked, the highest percent stated 4 5 their preference was the WEB programme. Those randomised to the conventional PR 6 7 8 classes who were less disabled (MRC 2) and younger would have preferred the web 9 10 programme, whereas the older patients preferred to attend the classes. Studies have shown 11 12 that most patients with coronary heart disease who are still working prefer to follow a home 13 14 15 based rehabFor program instead peer of conventional review supervised classes only (38). 16 17 Web based rehabilitation may inform the design of future trials. The data collected in this 18 19 20 study appears to have a role in the delivery of PR. However, a stratified approach may be 21 22 needed based upon patient need and choice of delivery to achieve the best outcomes for 23 24 patients and deliver a cost effective model of rehabilitation for a wider population. 25 26 27 Acknowledgements 28 This article presents independent research funded by the National Institute for Health 29 Research (NIHR) under its Research for Patient Benefit (RfPB) Programme (Grant Reference 30 Number PB-PG-0711-25127). The research took place at the University Hospitals of Leicester 31 NHS Trust and was supported by the NIHR Leicester Respiratory Biomedical Research Unit. 32 33 The views expressed are those of the author(s) and not necessarily those of the NHS, the http://bmjopen.bmj.com/ 34 NIHR or the Department of Health. 35 We thank Aga Glab, Chris Brough and Kelly Edwards for their involvement and support in 36 the study. 37 38 Contributors: 39 40 All authors of the paper have contributed to the design of the work, acquisition, analysis & 41 interpretation of the data. SS, JW, KE, EC, LA, CB, JB and SB were involved in the 42 development of the intervention and design of the trial. EC and SS have been involved in on September 23, 2021 by guest. Protected copyright. 43 drafting the work or revising it critically for important intellectual content and have given 44 the final approval of the version published. 45 46 47 Funding 48 This work was funded by the Research for Patient Benefit (RFPB)(PB-PG-0711-25127) which 49 is part of the funding body National Institute for Health Research (NIHR). The role of the 50 funder did not involve the study design; collection, management, analysis and interpretation 51 of data; writing and submission of the report. 52 53 54 Sponsor 55 The trial is sponsored by University Hospitals of Leicester NHS Trust Research and 56 Development department. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 Competing Interests: 4 None 5 6 7 Ethical approval 8 Ethics approval has been received from the Northampton Research Ethics Committee of the 9 UK National Research Ethics Service (Ethics Ref: 12/EM/0351). 10 11 Data sharing statement 12 Additional unpublished data from the study is still being collected and analysed and is only 13 14 available to members of the study team. 15 For peer review only 16 REDCAP 17 Study data were collected and managed using REDCap electronic data capture tools hosted 18 at UHL NHS Trust. REDCap (Research Electronic Data Capture) is a secure, web-based 19 20 application designed to support data capture for research studies, providing 1) an intuitive 21 interface for validated data entry; 2) audit trails for tracking data manipulation and export 22 procedures; 3) automated export procedures for seamless data downloads to common 23 statistical packages; and 4) procedures for importing data from external sources. 24 25 26 27 References 28 1.National Institute for Clinical Excellence. Clinical Guideline 101: Chronic obstructive 29 pulmonary disease – Management of chronic obstructive pulmonary disease in adults in 30 primary and secondary care, 21 June 2010. Ref Type: Generic 31 32 2. Bolton C, Bevan-Smith E, Blakey J, Crave P, Elkin S, Garrod R et al. BTS guidelines on 33 Pulmonary Rehabilitation in Adults. Thorax.(2013)68:ii1–ii30. http://bmjopen.bmj.com/ 34 35 36 3.Keating, A, Lee, A, Holland, A. What prevents people with chronic obstructive pulmonary 37 disease from attending pulmonary rehabilitation? A systematic review. Chronic Respiratory 38 Disease (2011) 8; 2: 89-99. 39 40 4. NHS England: Five Year Forward View. October 2014.Available at 41 42 https://www.england.nhs.uk/ourwork/futurenhs/ on September 23, 2021 by guest. Protected copyright. 43 44 5. Glasgow, R, Strycker, L, Kurz, D, Faber, A, Bell, H et al. Recruitment for an internet-based 45 diabetes self-management program: scientific and ethical implications. 46 Ann.Behav.Med.(2010) 40;1: 40-48. 47 48 49 6. Hoffmann, T, Russell, T, Thompson, L, Vincent, A, Nelson, M. Using the Internet to assess 50 activities of daily living and hand function in people with Parkinson's disease. 51 NeuroRehabilitation.(2008) 23;3 : 253-61. 52 53 7. Allen, M, Iezzoni, L, Huang, A, Huang, L, Leveille, S. Improving patient-clinician 54 55 communication about chronic conditions: description of an internet-based nurse E-coach 56 intervention. Nurs.Res.(2008) 57;2 : 107-12. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 28 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 8. Hurkmans, E, van den Berg, M, Ronday, K, Peeters, A, le Cessie, S et al. Maintenance of 4 physical activity after Internet-based physical activity interventions in patients with 5 rheumatoid arthritis. Rheumatology.(2010) 49;1 : 167-72. 6 7 8 9. Cruz-Correia, R, Fonseca, J, Amaro, M, Lima, L, Araujo, L et al. Web-based or paper-based 9 self-management tools for asthma--patients' opinions and quality of data in a randomized 10 crossover study. Stud.Health Technol.Inform.(2007) 127 :178-89. 11 12 10. Berman, R, Iris, M, Bode, R, Drengenberg, C. The effectiveness of an online mind-body 13 14 intervention for older adults with chronic pain. J.Pain (2009) 10;1 : 68-79. 15 For peer review only 16 11. Walker, E, Wexler,B, DiIorio, C, Escoffery, C, McCarty, F, Yeager, K. Content and 17 characteristics of goals created during a self-management intervention for people with 18 epilepsy. J.Neurosci.Nurs.(2009) 41;6 : 312-21. 19 20 21 12. Smeets, T, Brug, J, de Vries, H. Effects of tailoring health messages on physical activity. 22 Health Education Research. (2008) 23;3: 402-413. 23 24 13. Smit, ES, Evers, SM, de Vries, H, Hoving, C. Cost-effectiveness and cost utility of Internet- 25 based computer tailoring for smoking cessation. J Med Internet Res. (2013) 15;3: e57. 26 27 28 14.British Association for Cardiac Prevention and Rehabilitation. Standards and Core 29 components for Cardiovascular Disease Prevention and Rehabilitation; 2012. Available at 30 http://www.bacpr.com 31 32 33 15. Brough C, Boyce S, Houchen-Wolloff L, Sewell L, Singh S. J Evaluating the Interactive http://bmjopen.bmj.com/ 34 Web-Based Program, Activate Your Heart, for Cardiac Rehabilitation Patients: A Pilot Study. 35 J. Med Internet Res. (2014)16;10:e242 36 37 38 16.Devi R, Powell J, Singh S. A Web-Based Program Improves Physical Activity Outcomes in a 39 Primary Care Angina Population: Randomized Controlled Trial. J Med Internet Res (2014) 40 16(9):e186. DOI: 10.2196/jmir.3340 41 on September 23, 2021 by guest. Protected copyright. 42 17.Mitchell, K, Johnson-Warrington, V, Apps, L, Bankart, J, Sewell, L, Williams, J et al. A self- 43 44 management programme for COPD: a randomised controlled trial ERJ (2014) published on 45 line ahead of print doi: 10.1183/09031936.00047814 46 47 18.Apps, L, Mitchell, K, Harrison, S, Sewell, L, Williams, J, Young, H, Steiner, M, Morgan, M, 48 Singh, S.The development and pilot testing of the Self-management Programme of Activity, 49 Coping and Education for Chronic Obstructive Pulmonary Disease (SPACE for COPD). Int. J of 50 51 COPD (2013) 8; 317-27 52 53 19. Fletcher, C. Standardised questionnaire on respiratory symptoms: a statement prepared 54 and approved by the MRC committee on the aetiology of chronic bronchitis (MRC 55 breathlessness score). Br Med J. (1960) 2:1665. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 20. Singh, S, Morgan, M, Scott, S, Walters, D, Hardman, A. Development of a shuttle walking 4 test of disability in patients with chronic airways obstruction. Thorax (1992) 47;12 : 1019-24. 5 6 7 21. Revill, S, Morgan, M, Singh, S, Williams, J, Hardman, A. The endurance shuttle walk: a 8 new field test for the assessment of endurance capacity in chronic obstructive pulmonary 9 disease. Thorax (1999) 54;3 : 213-22. 10 11 22. Williams, J, Singh, S, Sewell, L, Guyatt, G, Morgan, M. Development of a self-reported 12 Chronic Respiratory Questionnaire (CRQ-SR). Thorax (2001) 56;12 : 954-59. 13 14 15 23. Zigmond,For A. S. and R.peer P. Snaith. The hospitalreview anxiety and depression only scale.Acta 16 Psychiatr.Scand. (1983) 67;6 : 361-70. 17 18 24. Jones, P, Harding, G, Berry, P, Wiklund, I, Chen, W-H, Kline Leidy, N. Development and 19 20 first validation of the COPD Assessment Test Eur Respir J (2009) 34:648-654. 21 22 25. Vincent E, Sewell, L., Wagg K, Deacon S, Williams, J., and Singh, S. J. Measuring a change 23 in self efficacy following Pulmonary Rehabilitation: An evaluation of the PRAISE tool. Chest 24 (2011) 140;6 1534-1539. 25 26 27 26. White, R, Walker, P, Roberts, S, Kalisky, S, White, P. Bristol COPD Knowledge 28 Questionnaire (BCKQ): testing what we teach patients about COPD. Chron.Respir.Dis.(2006) 29 3;3 : 123-31. 30 31 27. Rutten-van Molken, M, Oostenbrink, J, Tashkin, D, Burkhart, D, Monz, B. Does quality of 32 33 life of COPD patients as measured by the generic EuroQol five-dimension questionnaire http://bmjopen.bmj.com/ 34 differentiate between COPD severity stages? Chest (2006) 130;4 : 1117-28. 35 36 28. Thompson S and Wordsworth S. An annotated cost questionnaire for completion by 37 patients. UK working party on patient costs. 1-3-2001. HERU Discussion Paper 03/01. 38 39 40 29. Chaplin, E, Hewitt, S, Apps, L, Edwards, K, Brough, C, Glab, A, Bankart, J, Jacobs, R, Boyce, 41 S, Williams, J, Singh, S. The Evaluation of an interactive web-based Pulmonary 42 Rehabilitation programme: protocol for the WEB SPACE for COPD feasibility study. BMJ on September 23, 2021 by guest. Protected copyright. 43 open (2015) 1-7 44 45 46 30.Steiner M, Holzhauer-Barrie J, Lowe D, Searle L, Skipper E, Welham S, Roberts CM. 47 Pulmonary Rehabilitation: Steps to breathe better. National Chronic Obstructive Pulmonary 48 Disease (COPD) Audit Programme: Clinical audit of Pulmonary Rehabilitation services in 49 England and Wales 2015. London: RCP, 2016. 50 www.rcplondon.ac.uk/projects/outputs/pulmonary-rehabilitation-steps-breathe-better 51 52 53 31.Voncken-Brewster, V, Tange,H, de Vries, H, Nagykaldi, Z, Winkens, B, van der Weijden, T. 54 A randomised controlled trial evaluating the effectiveness of a web-based, computer- 55 tailored self-management intervention for people with or at risk for COPD. Int. J of COPD 56 (2015) 10: 1061-1073. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 28 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 32.Pinnock, H, Hanley J, McCloughan L, Todd A, Krishan A, Lewis S et al. Effectiveness of 4 telemonitoring integrated into existing clinical services on hospital admission for 5 exacerbation of chronic obstructive pulmonary disease: researcher blind, multicentre, 6 7 randomised controlled trial. BMJ (2013) 347:1-16. 8 9 33. Singh S, Jones P , Evans R , Morgan M . Minimum clinically important improvement for 10 the incremental shuttle walking test. Thorax (2008) 63 (9): 775-777 11 12 34.Murray E. Web-based interventions for behaviour change and self-management: 13 14 potential, pitfalls, and progress. Med 2.0 (2012)1(2):e3. 15 For peer review only 16 35.Vandelanotte C, Dwyer T, Van Itallie A, Hanley C, Mummery WK. The development of an 17 internet-based outpatient cardiac rehabilitation intervention: a Delphi study. BMC 18 Cardiovasc Disord (2010)10:27. 19 20 21 36. Gibb, M, Willott, V, Lohar, S, Ward, S, Bolton C , McAlinden, P et al. An evaluation to 22 understand the use of technology within a COPD population. Thorax (2013) 68:A96-A97. 23 24 25 37.Neville LM, O'Hara B, Milat A. Computer-tailored physical activity behavior change 26 27 interventions targeting adults: a systematic review. Int J Behav Nutr Phys Act (2009)6:30 28 29 38.Grace SL, McDonald J, Fishman D, Caruso V. Patient preferences for home-based versus 30 hospital-based cardiac rehabilitation. J Cardiopulm Rehabil (2005) 25;1:24-29. 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 23, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21

22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34

35 36 37 38 39 40 41 42 on September 23, 2021 by guest. Protected copyright. 43 44 45 46 47 Figure 1 SPACEforCOPD dashboard screen showing tasks completed in stage 1 as well as an 48 49 overview of exercise progression, goals, knowledge and symptom diary. 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 28 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 4 2646 Patients invited to 5 participate 6 7 Excluded as not Eligibility 8 eligible (n=244) 9 MRC <2 = 3 10 No internet access = 111

641 Patients 11 Co morbidities = 1 12 Preference for PR classes 13 = 54 14 Preference WEB = 2 294 were not 15 interestedFor peer review onlyOn other study = 8 16 Recruitment Not COPD = 42 17 PR in <12/12 = 3 Unwell = 2 18 Consented & Randomised No time = 1 19 (n=103) RIP = 1 20 UTC = 9 21 Unknown = 7 22 23 24 25 Allocation 26 27

28 Usual Care - PR WEB (n=51)

29 (n=52) 30 31 WEB Introduction 32 33 http://bmjopen.bmj.com/

34 Withdrawals (n= 12) 35 Co morbidities = 2 Weekly phone call or emails 36 Social / family reasons = 1 37 DNA classes = 6 No time to commit to 38 Withdrawals (n= 29) study = 1 39 Unable to contact = 4 Intervention Does not want to exercise = 1 40 Exercises at home = 1 Time = 5 41 Preference PR classes = 3 Co-morbidities = 5 on September 23, 2021 by guest. Protected copyright. 42 COPD ill health = 5 43 Could not engage with website 44 = 3 45 Not suitable for home exercise 46 = 2 Broken computer = 1 47 Family circumstances = 1 48 49 50

51 Discharge Completers (n=40) Completers (n=22) 52 53 54 55 Figure 2 Consolidation Standards of Reporting Trials flow diagram of participation. MRC: 56 Medical Research Council; COPD: Chronic Obstructive Pulmonary Disease; PR: Pulmonary 57 Rehabilitation. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 4 PR WEB 5 Age (Years) 66.1 ± 8.1 66.4 ± 10.1 6 7 Gender (% male) 63.5 74.5 8 FEV1 (% predicted) 55.0 ± 20.5 58.7 ± 29.1 9 BMI (kg/m2) 29.3 ± 6.3 27.9 ± 6.4 10 11 12 MRC (IQR) 3 (2-4) 3 (2-4) 13 MRC (n) 14 2 21 14 15 3 For peer review13 only20 16 4 14 12 17 5 1 3 18 19 20 Baseline ISWT (m) 284.2 ± 156.0 296.7 ± 180.8 21 Baseline ESWT (secs) 246.2 ± 144.0 241.7 ± 209.7 22 Pre CRQ SR -D 2.7 ± 1.1 2.7 ± 1.2 23 24 Pre CAT 20.8 ± 7.5 20.8 ± 8.6 25 Pre PRAISE 45.7 ± 7.7 45.6 ± 7.7 26 Pre HADS 27 Anxiety 7.1 ± 5.0 7.9 ± 4.8 28 Depression 5.8 ± 3.6 6.4 ± 3.8 29 30 Pre BCKQ 37.1 ± 12.5 33.9 ± 8.6 31 32 Table 1 Baseline characteristics. Data are presented as n or mean ± SD. FEV1: forced 33

expiratory volume in 1 s; BMI: body mass index; MRC: Medical Research Council; ISWT: http://bmjopen.bmj.com/ 34 Incremental Shuttle Walk Test; ESWT: Endurance Shuttle Walk Test; CRQ-SR- D: Chronic 35 Respiratory disease Questionnaire – Self Report - Dyspnoea; CAT: COPD Assessment Tool; 36 37 PRAISE: PR Adapted Index of Self Efficacy; HADS: Hospital Anxiety and Depression Scale; 38 BCKQ: Bristol COPD Knowledge Questionnaire. 39 40 41

PR WEB Between group on September 23, 2021 by guest. Protected copyright. 42 43 differences 44 (p value) 45 Change in ISWT 12.9 ± 58.6 10.9 ± 46.4 0.9 46 47 Change in ESWT 184.5 ± 247.4 189.0 ± 211.1 0.9 48 49 50 Change in CRQ-D 0.7 ± 1.2 0.8 ± 1.0 0.7 51 52 53 Table 2 Clinical outcome measures. Data are presented as mean ± SD. ISWT: Incremental 54 55 Shuttle Walk Test; ESWT: Endurance Shuttle Walk Test; CRQ-SR- D: Chronic Respiratory 56 disease Questionnaire – Self Report – Dyspnoea 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 28 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 4 5 6 7 8 *p < 0.001 9 *p < 0.001 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 Figure 3 Exercise capacity. Within group changes of the ESWT: endurance shuttle walk test. 39 40 41 on September 23, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 4 *p < 0.001 5 *p < 0.001 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Figure 4 Quality of life. Within group changes of CRQ-D: Chronic Respiratory disease http://bmjopen.bmj.com/ 34 Questionnaire – Self Report – Dyspnoea 35 36 37 38 39 40 41 42 on September 23, 2021 by guest. Protected copyright. 43 44 Stage No. of participants 45 No WEB introduction completed 5 46 Not registered 7 47 48 Stage 1 4 49 Stage 2 11 50 Stage 3 2 51 Stage 4 0 52 53 54 Table 3 Drop out stages of the WEB programme. 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 28 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 4 5 WEB PATIENTS PR PATIENTS 6 7 WEB CLASSES WEB CLASSES 8 9 NO PREFERENCE NOT STATED NO PREFERENCE NOT STATED 10 11 12 20% 15% 13 14 37% 39% 15 For peer review15% only 16 17 27% 18 19 16% 31% 20 21 22 23 Fig 5a and b Patient preference 24 25 26 27 28 29 30 31 Intervention Standard care Mean difference 32 33 Mean (SD) Mean (SD) http://bmjopen.bmj.com/ 34 35 Baseline 0.225(0.609) 0.261(0.577) -0.035 36 37 Follow-up visit 0.329(0.566) 0.313(0.580) 0.016 38 39 40 Table 4: EQ-5D score by group 41 42 on September 23, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 Supplement for the Web paper 4 5 Stages of the WEBsite 6 7 8 9 SPACEforCOPD® (Figure 1) is an interactive Web-based PR programme that offers a 10 comprehensive package of exercise and education. It was developed following the 11 success of Activate Your Heart which was devised by cardiac rehabilitation specialists 12 at our institution, UHL NHS trust. The programme is password protected; each 13 14 participant was given their own unique password to access the SPACEforCOPD® 15 programme.For All participants peer were ablereview to record and monitor only their exercises and 16 strength training as well as interacting with members of the research team (Figure 17 2). 18 19 The SPACEforCOPD® programme was structured to guide the user through four 20 21 stages that each have specific tasks the user needs to achieve before progressing 22 onto the next stage (Figure 3). Tasks included creating and updating their own short- 23 term goals, completing knowledge tests on COPD and exercising safely, and reading 24 specific topics such as inhaler techniques. The educational reading material includes 25 videos and covers topics such as disease education, managing breathlessness, 26 27 recognising the signs and symptoms of an exacerbation, energy conservation and 28 diet and healthy eating (Figure 4). 29 30 In Stage 1, participants were asked to do a multiple choice questionnaire to establish 31 their knowledge regarding the principles of exercising safely. A score of 80% was set 32 as a threshold to ensure understanding of these principles. In Stages 2-4, participants 33 http://bmjopen.bmj.com/ 34 were required to record all their exercises, both aerobic and strength, in an exercise 35 diary. During Stage 2, participants were advised to record their aerobic exercise, 5 36 days out of 7. The intensity of the exercise was based on their performance on the 37 baseline ISWT and ESWT exercise tests and prescribed at 85% of baseline 38 performance. In Stage 3, as well as completing 5 out of 7 days of aerobic exercise, 39 patients were asked to also start their resistance training, completing 3 strength 40 41 sessions in a week. Finally, in Stage 4, participants were required to maintain both 42 the aerobic and strength components of the exercise programme for a further 2 on September 23, 2021 by guest. Protected copyright. 43 weeks. There was also interactivity around stress management, knowledge and 44 smoking cessation, if appropriate. For smokers, a cost calculator was developed that 45 would calculate how much the user had spent or saved since starting the 46 47 programme. This was delivered along with advice and support to stop smoking. 48 49 Other features embedded within the SPACEforCOPD® programme included a forum 50 where patients were also able to share views and experiences with other 51 programme users, a blog, and a frequently asked questions section. The forum was 52 monitored and moderated, as necessary, by the research team. In addition patients 53 54 were able to communicate privately with the research team via the Ask the Expert 55 messaging facility. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 28 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 The research team members were provided with individual passwords to access the 4 administration section of the programme; this allowed them to view and monitor 5 individual patients’ progress and view patient login data. Patients’ were contacted 6 7 once a week via email or phone to discuss how they were progressing, any issues or 8 barriers to them not continuing the programme. 9 10 All data captured on the programme were encrypted to safeguard patient 11 confidentiality. 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 Figure 1. SPACE for COPD website homepage 41

on September 23, 2021 by guest. Protected copyright. 42

43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 Figure 2. Recording details of exercise and strength training 32 33 http://bmjopen.bmj.com/ 34 35

36 37 38 39 40 41 42 on September 23, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55

56 57 58 Figure 3. Part of the dashboard showing what tasks need to be completed in Stage 2 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 28 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 Figure 4. Reading material section 26 27 28 29 30 31 32 33

http://bmjopen.bmj.com/ 34 35 36

37 38 39 40 41 42 on September 23, 2021 by guest. Protected copyright. 43 44 45 46

47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

An interactive web-based Pulmonary Rehabilitation programme: A randomised controlled feasibility trial.

ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2016-013682.R1

Article Type: Research

Date Submitted by the Author: 10-Nov-2016

Complete List of Authors: Chaplin, Emma; Centre for Exercise and Rehabilitation Science, Leicester Respiratory Biomedical Research Unit Hewitt, Stacey; Centre for Exercise and Rehabilitation Science, Leicester Respiratory Biomedical Research Unit Apps, Lindsay; Centre for Exercise and Rehabilitation Science, Leicester Respiratory Biomedical Research Unit Bankart, M. John; Keele University, Institute of Primary Care Pulikottil-Jacob, Ruth; Warwick Medical School , Warwick Evidence Boyce, Sally; Centre for Exercise and Rehabilitation Science, Leicester Respiratory Biomedical Research Unit Morgan, Mike; University Hospitals of Leicester NHS Trust, Respiratory Medicine Williams, Johanna; Glenfield Hospital, University Hospitals of Leicester NHS Trust, Respirartory Medicine Singh, Sally; University Hospitals of Leicester NHS Trust, http://bmjopen.bmj.com/ Cardiac/Pulmonary Rehabilitation

Primary Subject Rehabilitation medicine Heading:

Secondary Subject Heading: Respiratory medicine

SPACE for COPD, chronic obstructive pulmonary disease, Internet, Web- Keywords:

based, pulmonary rehabilitation on September 23, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

23 24 Key words: SPACE for COPD; Internet; web-based; chronic obstructive pulmonary disease; 25 pulmonary rehabilitation 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 23, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 32 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 Abstract: 4 5 Objectives: The aim of this study was to determine if an interactive web-based Pulmonary 6 7 Rehabilitation (PR) programme is a feasible alternative to conventional PR. 8 9 Design: Randomised controlled feasibility trial 10 11 Setting: Participants with a diagnosis of COPD were recruited from PR assessments, primary 12 care and community rehabilitation programmes. Patients randomised to conventional 13 14 rehabilitation commenced the programme according to the standard care at their referred 15 site on the nextFor available peer date. review only 16 17 Participants: 103 patients were recruited to the study and randomised: 52 to conventional 18 rehabilitation [mean (±SD) age 66 (±8) years, MRC 3 (IQR2-4)]; 51 to the web arm [mean 19 20 (±SD) age 66 (±10) years, MRC 3 (IQR2-4)]. Participants had to be willing to participate in 21 either arm of the trial, have internet access and be web literate. 22 23 Interventions: Patients randomised to the web-based programme worked through the 24 website, exercising and recording their progress as well as reading educational material. 25 Conventional PR consisted of twice weekly, two hourly sessions (an hour for exercise 26 27 training and an hour for education). 28 29 Outcome measures: Recruitment rates, eligibility, patient preference and drop-out and 30 completion rates for both programmes were collected. Standard outcomes for a PR 31 assessment including measures of exercise capacity and quality of life questionnaires were 32 33 also evaluated. http://bmjopen.bmj.com/ 34 35 Results: A statistically significant improvement (p≤ 0.01) was observed within each group in 36 the ESWT (WEB: mean change 189 ± 211.1; PR classes: mean change 184.5 ± 247.4secs) and 37 CRQ-D (WEB: mean change 0.7 ± 1.2; PR classes: mean change 0.8 ± 1.0). However there 38 were no significant differences between the groups in any outcome. Dropout rates were 39 40 higher in the web programme (57% v 23%). 41 42 Conclusion: An interactive web-based PR programme is feasible and acceptable when on September 23, 2021 by guest. Protected copyright. 43 compared to conventional PR. Future trials maybe around choice based PR programmes for 44 select patients enabling stratification of patient care. 45 46 47 Trial registration number ISRCTN03142263 48 49 50 Strengths and Limitations of this study 51 • The study concentrates on the feasibility of an interactive web-based Pulmonary 52 53 Rehabilitation programme (SPACE: Self-management Program of Activity, Coping 54 and Education for COPD®). 55 • It provides data on recruitment, eligibility and patient preference which will inform 56 future trials around choice based programmes for select patients. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 • A limitation to the study was a lack of engagement despite patient involvement in 4 the site development. 5 Limitations were identified when recruiting patients to a technology based 6 • 7 intervention, in that patients needed to be competent users with an in-depth, 8 specific web-based knowledge. 9 • The study compares a variety of clinical outcomes between a web-based and a 10 conventional rehabilitation PR programme. This facilitates a personalised approach 11 to rehabilitation. 12 13 14 15 IntroductionFor peer review only 16 17 Chronic Obstructive Pulmonary Disease is the fourth leading cause of death in the United 18 19 20 Kingdom (UK) and is characterised by a progressive deterioration of debilitating symptoms 21 22 and increasingly frequent exacerbations. Pulmonary rehabilitation (PR) has been proven to 23 24 be effective in improving quality of life, psychological functioning and physical activity and 25 26 27 national guidelines recommend that PR should be offered and made available to all those 28 29 with COPD (1). The standard provision of PR is a supervised package of exercise and 30 31 education usually twice a week for a minimum of six weeks, which is either hospital or 32 33 http://bmjopen.bmj.com/ 34 community based, and supported by a home exercise programme (2). However the barriers 35 36 to uptake of a PR programme have previously been reported (3) which included transport, 37 38 39 the perceived benefits of PR, disruption to usual routine and the timings of programmes. 40 41 These factors play some contribution as to why programmes have poor attendance and 42 on September 23, 2021 by guest. Protected copyright. 43 adherence resulting in many of the programmes reporting dropout rates as high as 50%. 44 45 46 47 Ongoing changes and challenges means that the NHS and the services it provides need to 48 49 adapt to take advantage of and capitalise on the opportunities that new technologies and 50 51 treatments can offer to patients (4). There is a growing evidence base for the use of the 52 53 54 internet in the management of many chronic conditions in areas as diverse as the 55 56 management of diabetes, Parkinson’s disease, depression, rheumatoid arthritis, asthma, 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 32 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 chronic pain and epilepsy (5-11). Computer-tailored interventions have been shown to 4 5 effectively improve health behaviours such as physical activity (12) and be cost effective 6 7 8 (13). The provision of a menu based and patient centred service is said to be essential to 9 10 improve uptake and completion rates within Cardiac Rehabilitation (CR) (14). However at 11 12 present there is no choice within Pulmonary Rehabilitation. A web-based PR programme has 13 14 15 the potentialFor to be a novel peer and effective reviewapproach to increasing only patient choice in the mode 16 17 of delivery and setting of rehabilitation (especially to those patients who decline the offer of 18 19 20 conventional pulmonary rehabilitation) while simultaneously increasing the capacity of PR. 21 22 23 We have previously developed and described ‘ACTIVATE YOUR HEART®’ (AYH)(15; 16) which 24 25 is an interactive web-based cardiac rehabilitation (CR) programme that has proved very 26 27 28 popular with patients. Brough et al (2014) reported a significant improvement in exercise 29 30 capacity and quality of life in patients that completed the web-based programme. Following 31 32 the success of AYH, we have developed a prototype website based on the educational 33 http://bmjopen.bmj.com/ 34 35 content of the ‘SPACE for COPD®’ self-management workbook. SPACE (Self-management 36 37 programme of Activity, Coping and Education) for COPD® (17) is a structured programme of 38 39 40 exercise, education and psychosocial support which has been developed by our institution 41 42 as a collaboration between experts, patients and carers and has been awarded a Crystal on September 23, 2021 by guest. Protected copyright. 43 44 Mark for Clarity by the Plain English Campaign (18). 45 46 47 48 The aim of this feasibility study was to provide quantitative, economic and technical data to 49 50 see if an interactive web-based PR programme was a feasible alternative compared with 51 52 conventional PR. This included: 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 1. Gathering information regarding the recruitment rate of patients who were eligible and 4 5 willing to be randomised to either the web-based or the conventional rehabilitation 6 7 8 programme, and to monitor retention and drop out through all stages of the programmes. 9 10 2. Comparing a variety of clinical outcomes between a web-based and a conventional 11 12 rehabilitation PR programme in order to test out the various components of the 13 14 15 interventionFor and identify peer any technical orreview other difficulties that only may be inherent in the 16 17 delivery of a web-based PR programme. 18 19

20 21 Methods 22 23 Participants 24 25 Eligibility criteria for participants 26 27 28 Eligible participants had an established diagnosis of Chronic Obstructive Pulmonary Disease 29 30 (COPD) defined as a forced expiratory volume in one second (FEV1),post-bronchodilation of 31 32 <80% and a predicted ratio of FEV1 to forced vital capacity of 0.70 and a Medical Research 33 http://bmjopen.bmj.com/ 34 35 Council (MRC) dyspnoea score (19) of between 2 and 5. Patients had to be willing to partake 36 37 in either arm of the study. Access to the internet for more than 3 months, the ability to 38 39 40 navigate around a variety of websites (for example, uses online shopping or banking 41 42 websites) and regular use of email was required. Patients also had to be able to read and on September 23, 2021 by guest. Protected copyright. 43 44 write in English. 45 46 47 48 49 Patients were excluded if they were unable to participate in the exercise component of the 50 51 rehabilitation programme due to other comorbidities or had done PR in the previous 12 52 53 54 months. Eligible patients had to be willing and able to take part in the web-based 55 56 programme. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 32 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 Setting 4 5 6 Participants to the study were primarily recruited from those patients that had been 7 8 9 referred for pulmonary rehabilitation at University Hospitals of Leicester (UHL) NHS Trust. 10 11 Recruitment was also directly from primary care and community rehabilitation services 12 13 within Leicester Partnership Trust (LPT) and eligible participants were identified from the 14 15 For peer review only 16 research participant database of the Leicester Respiratory Biomedical Research Unit and 17 18 Pulmonary Rehabilitation Department. 19 20 21 22 23 Randomisation 24 25 Patients were randomised to either the conventional rehabilitation programme as is 26 27 28 standard at their referred site or the web-based Pulmonary Rehabilitation programme 29 30 (SPACE for COPD®). Randomisation to the treatment group allocation was on a 1:1 ratio to 31 32 either group and was performed using a web based programme 33 http://bmjopen.bmj.com/ 34 35 (www.sealedenvelope.com). 36 37 38 39 Trial Interventions 40 41 42 Intervention group – web-based pulmonary rehabilitation programme on September 23, 2021 by guest. Protected copyright. 43 44 Following randomisation to the intervention group, patients attended a standardised 45 46 introductory session where participants were given a password protected secure log-in to 47 48 49 the website as well as written instructions on website navigation. 50 51 52 53 54 Patients were directed to all the relevant sections on the website including the home 55 56 exercise programme and goal setting. There was also an individualised webpage (Figure 1) 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 featuring a personalised action plan designed to assist in the management of exacerbations 4 5 which was completed by the rehabilitation specialist in conjunction with the patient. 6 7 8 9 10 As in conventional pulmonary rehabilitation, patients were encouraged to exercise on a 11 12 daily basis at home and record their progress in the online exercise diary section. The 13 14 15 exercise programmeFor consisted peer of both aerobicreview and strength training. only The intensity of the 16 17 walking was based on their performance on the baseline maximal shuttle walking exercise 18 19 20 tests and prescribed at 85% of baseline performance. An exercise target was set by the 21 22 patient to achieve and work towards each week. Strength training consisted of both upper 23 24 and lower limb resistance training with hand held weights. Patients recorded how difficult 25 26 27 they found the walking and strength training using a visual analogue scale (VAS). Both the 28 29 walking time and strength were progressed maintaining a VAS rating of 4-7. Throughout the 30 31 duration of the web-based programme the patient’s progress was reviewed online and 32 33 http://bmjopen.bmj.com/ 34 there was weekly contact between the patient and the rehabilitation specialist via email or 35 36 telephone using a standardised proforma. Motivational interviewing techniques were used 37 38 by the healthcare professional to ensure that patients were helped to progress their 39 40 41 exercise programme in both the aerobic and strength training appropriately and to answer 42 on September 23, 2021 by guest. Protected copyright. 43 any queries that arose. 44 45 46 47 48 The educational content of the web-based programme was based on the ‘SPACE for COPD®’ 49 50 manual. Patients worked through the website content at their own pace, however certain 51 52 53 milestones needed to be completed or achieved before further content could be accessed in 54 55 order to ensure appropriate progress through the programme (see supplement for 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 32 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 description of the different stages on the website). It was anticipated from previous work 4 5 (14) that it would take approximately 6-8 weeks to work through the online programme. 6 7 8 9 10 Standard care group - conventional pulmonary rehabilitation programme 11 12 Patients randomised to standard care commenced conventional rehabilitation according to 13 14 15 the standardFor care at their peer referred site whichreview was either hospital only or community based. The 16 17 hospital programme consisted of 7 weeks (4 weeks supervised; 3 weeks unsupervised) in 18 19 20 total. Patients were advised to not attend if they were having an exacerbation. Any sessions 21 22 missed could be completed later due to it being a rolling programme. In the community 23 24 based programmes, patients could attend a maximum of 12 sessions within the closed 25 26 27 programme. 28 29 Conventional pulmonary rehabilitation programmes at either referral site consisted of twice 30 31 weekly sessions each lasting two hours which were divided into an hour for exercise training 32 33 http://bmjopen.bmj.com/ 34 and an hour for an education session covering a variety of relevant self-management topics. 35 36 The exercise training consisted of both aerobic and resistance training. A training walking 37 38 speed was prescribed from the ISWT (20) and ESWT (21) performed at baseline. Walking 39 40 41 time was progressed maintaining a moderate to severe breathlessness as defined by the 42 on September 23, 2021 by guest. Protected copyright. 43 BORG dyspnoea scale (22). Patients were instructed to walk daily at their PR class training 44 45 46 speed. Strength training consisted of both upper and lower limb resistance training with 47 48 dumbbells which was based on 1 repetition maximum (1-RM). Progression was achieved by 49 50 maintaining a BORG perceived exertion (23) rating of 13 to 15. Static cycling was completed, 51 52 53 if tolerated, and intensity was prescribed on the basis of the patient’s breathlessness and 54 55 perceived exertion symptom scores. Patients were encouraged to also complete a home 56 57 exercise programme on the days when they did not attend rehabilitation classes and to fill 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 in an exercise diary. This enabled the patients’ progress to be monitored. The educational 4 5 sessions were conducted as group sessions and delivered by experts in their field. Topics 6 7 8 included medication, relaxation skills, chest clearance and breathlessness management and 9 10 energy conservation. 11 12 13 14 15 Outcome measuresFor peer review only 16 17 18 All the measures used and collected in the trial including clinical (the incremental (ISWT) and 19 20 21 endurance (ESWT) shuttle walk tests; chronic respiratory questionnaire- self reported (CRQ- 22 23 SR)(24), Hospital Anxiety and Depression Scale (HADS)(25), COPD Assessment Tool 24 25 (CAT)(26), PR Adapted Index of Self-Efficacy (PRAISE)(27), Bristol COPD Knowledge 26 27 28 questionnaire (BCKQ)(28), Euro-QOL (EQ-5D-5L)(29), patient cost questionnaire (30)) and 29 30 non-clinical have previously been described in the study protocol (31). Clinical measures 31 32 33 were performed at baseline and repeated again at the discharge assessment following http://bmjopen.bmj.com/ 34 35 completion of either rehabilitation programme (usually approximately 6-7 weeks after 36 37 starting the programme) and were conducted by a research physiotherapist who was 38 39 40 blinded to treatment group allocation. Patients were classed as a completer if they had 41 42 reached stage 3 or above of the web programme, achieving 75% of the programme which is on September 23, 2021 by guest. Protected copyright. 43 44 standard in clinical practice for those attending classes. Those patients randomised to the 45 46 47 website were offered conventional PR classes if they felt it would be more beneficial at 48 49 discharge. 50 51 Non-clinical outcomes included a web-usage audit for the internet-based programme, 52 53 54 recruitment rates, eligibility and patient preference as well as drop out and completion 55 56 rates in both treatment groups. Any serious adverse events were reported to the sponsor. A 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 32 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 serious adverse event was defined as an acute exacerbation of their COPD that resulted in a 4 5 hospital admission. In order to assess the patients’ ability to exercise safely, an exercise 6 7 8 safety quiz was completed online before being able to progress onto stage 2 of the 9 10 programme which involved exercising. Patients were then monitored online and through 11 12 the weekly contacts. 13 14 15 Qualitative andFor physical activitypeer data are review to be presented in future only publications. 16 17 18 19 20 Quantitative Data Analysis 21 22 Data was entered and stored on a secure web-based system (REDCAP) which has 23 24 discrepancy management features. Data was then transferred from REDCAP to the 25 26 27 Statistical Package for the Social Sciences (SPSS) version 18 (SPSS Ltd, Woking,Surrey, UK). 28 29 The data were checked for normality before baseline characteristics were compared 30 31 between-groups using an independent t-test. Analysis was primarily descriptive i.e. 32 33 http://bmjopen.bmj.com/ 34 estimation of means and standard deviations, proportion of patients eligible/ willing to 35 36 participate in the study. A paired t-test was used to compare within-group changes and an 37 38 independent t-test was used to compare the differences between the two treatment groups 39 40 41 in the ISWT, ESWT and CRQ-D at the two different time points. 42 on September 23, 2021 by guest. Protected copyright. 43 44 Results 45 46 47 103 patients were recruited and randomised to the study between May 2013 and July 2015: 48 49 52 to the conventional PR group and 51 to the web group. Figure 2 shows the flow of 50 51 eligibility, screening, randomisation and follow-up in the study. No significant differences 52 53 54 between the groups’ baseline characteristics or outcome measures (Table 1) were seen. 55 56 More patients dropped out from the web intervention group (n = 29) but there were no 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 significant differences between the baseline characteristics of those patients that dropped 4 5 out of the 2 groups. Reasons for dropouts are listed in Figure 2. The only significant 6 7 8 characteristic between web completers and dropouts was the pre anxiety scores (p<0.05) 9 10 with those that dropped out being more anxious (Table 2). 11 12 Clinical Outcome measures 13 14 15 A statisticallyFor significant improvementpeer (p≤0.01)review was observed withinonly each group in the ESWT 16 17 (WEB: mean change 189 ± 211.1; PR classes: mean change 184.5 ± 247.4 secs) (Figure 3) and 18 19 20 CRQ-D (WEB: mean change 0.7 ± 1.2; PR classes: mean change 0.8 ± 1.0) (Figure 4). There 21 22 were no significant differences between the groups in any clinical outcome. All outcome 23 24 measures used were feasible to administer. 25 26 27 Non Clinical study outcomes 28 29 The average number of weeks to complete the website was 11 ± 4 with an average number 30 31 of 4 logins per week. Patients tended to spend the longest time in Stage 2. This was where 32 33 http://bmjopen.bmj.com/ 34 the exercise programme was started and had the most activities to complete before further 35 36 content could be accessed. The education material was tailored to the gaps in the patient’s 37 38 knowledge based on those identified from the BCKQ completed at registration and 39 40 41 therefore time was spent in different areas accordingly. The stage at which patients 42 on September 23, 2021 by guest. Protected copyright. 43 dropped out at are listed in Table 3. The majority of patients dropped out at the beginning 44 45 46 of the web programme which suggests once the patient was engaged with the programme, 47 48 they were able to complete it. Those that dropped out tended to be mostly MRC 3, had a 49 50 lower baseline ISWT and a significantly higher HADS anxiety score at baseline compared 51 52 53 with those that completed the web programme (Table 2). In those patients who had been 54 55 randomised to the PR classes, the main reasons that they were withdrawn, were the 56 57 patients not starting the classes and being unable to contact the patient (n=6). 25% of the 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 32 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 web withdrawals would have preferred to have attended the classes compared with 54% of 4 5 patients that attended conventional PR classes preferring to have done the web 6 7 8 programme. 9 10 When patients were asked their treatment preference prior to being randomised, the 11 12 largest proportion of patients wanted the web programme (n = 38%) (Figure 5a & b). Of the 13 14 15 22 patients thatFor completed peer the web programme, review only 3 patients only (n=14%) felt they would 16 17 like to attend the PR classes. 18 19 20 21 22 Discussion 23 24 Pulmonary Rehabilitation is one of the most effective non pharmaceutical therapies for 25 26 27 patients with COPD, which offers long term benefits. Issues such as transport and locality 28 29 still persist for many patients some of which could be addressed by improving accessibility. 30 31 Data from a recent national COPD audit (32) suggests that the capacity of PR programmes is 32 33 http://bmjopen.bmj.com/ 34 inadequate to meet the demand or need. Alternative or more flexible provisions of PR 35 36 programmes therefore need to be considered. 37 38 Voncken-Brewster et al (2015) (33) tested the effectiveness of a web-based computer- 39 40 41 tailored COPD self-management intervention on physical activity and smoking behaviour. 42 on September 23, 2021 by guest. Protected copyright. 43 There were no statistically significant effects on health-related behavioural or clinical 44 45 46 outcomes. They state this may have been attributed to the low exposure to the application 47 48 and engagement with the program has been shown to be crucial for the effectiveness of 49 50 computer tailored interventions (34, 35). Another randomised controlled trial by Pinnock et 51 52 53 al in 2013 (36) found telemonitoring to not reduce hospital admission or improve patients’ 54 55 quality of life. The data from this feasibility study aimed to evaluate the application of an 56 57 exercise intervention as well as promoting self-management. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 The data from this study suggests that an interactive web-based programme has the 4 5 potential to be a feasible and acceptable alternative when compared with conventional PR. 6 7 8 Although both groups improved in the ISWT, the change wasn’t significant and did not meet 9 10 the minimally clinical important difference (MCID) of 48m (37). The baseline ISWT scores in 11 12 both groups were higher in these patients compared with those seen normally in our clinical 13 14 15 service and thereforeFor may peer account for thereview small change in the onlyISWT score. More patients 16 17 were MRC 2 in the PR group than are normally referred to the PR service. The change in 18 19 20 ISWT may also have been affected due to the structure of the PR programme which was 7 21 22 weeks (4 weeks supervised; 3 weeks unsupervised) for the majority of the patients and did 23 24 not meet the BTS guidelines (2) of a minimum of 12 supervised sessions. These guidelines 25 26 27 were not published until after the trial had started. Similar improvements in time (3 28 29 minutes) were seen in the endurance shuttle walk tests for both groups. Health related 30 31 quality of life, measured by the CRQ-SR, appeared to improve significantly, both groups 32 33 http://bmjopen.bmj.com/ 34 exceeding the MCID for the CRQ-D of 0.5. High dropout rates in the web arm of the study 35 36 may have influenced the outcome measures of exercise capacity. The study would suggest, 37 38 like other rehabilitation studies that using a HRQoL measure is a feasible primary outcome 39 40 41 measure. However despite recruiting a large number of participants, the high dropout rate 42 on September 23, 2021 by guest. Protected copyright. 43 and challenges experienced around a technology based intervention based on the findings 44 45 46 of this study, would potentially make future non-inferiority trials using the CRQ harder; a 47 48 larger sample size of patients in each group would be required. The use of preference based 49 50 RCT’s could be more appealing to patients therefore improving recruitment and retention 51 52 53 rates. 54 55 Adherence to web-based programs can vary due to many reasons ranging from lack of time, 56 57 to refusing to complete the program. Several features have been identified that could help 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 32 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 to improve adherence to a web program: making the program tailored to the user and 4 5 interactive (38) as well as allowing users to set personal goals (39). 6 7 8 Initial withdrawals in this study appeared to have been at the exercise stages of the web 9 10 programme as there was a higher dropout at stage 2. This component of the web 11 12 programme was simplified based on participant feedback. Both the length of time to 13 14 15 complete stageFor 2 (aerobic peer training) and 3review (strength and aerobic only training) were shortened 16 17 and completion rates improved. Similar numbers withdrew after this modification, however 18 19 20 these were due to exacerbations, other problems from co-morbidities and technology 21 22 problems meaning patients didn’t even register or come for their introduction. Although 23 24 more patients withdrew from the web arm of the study (12 from the PR group v’s 29 from 25 26 27 the web group), there were no significant differences between the groups in any outcome. 28 29 Those that did complete the website, did as well as those that had completed conventional 30 31 PR. 32 33 http://bmjopen.bmj.com/ 34 Priorities around the use of technology within the NHS are changing (4) but this will not be 35 36 without its challenges. A previous evaluation to explore the use of technology within a 37 38 COPD population which was carried out by our institution found although patients owned a 39 40 41 computer or mobile phone, usage was limited and was predominantly within the younger 42 on September 23, 2021 by guest. Protected copyright. 43 age range (40). 17% of interested participants in this study did not have access to the 44 45 46 internet and highlights the need to assess access as well as the competency of patients 47 48 being able to use the web prior to starting the web programme. The majority of patients 49 50 that entered the study expressed a preference for randomisation to the website arm of the 51 52 53 study, showing that there is a desire for this type of intervention. Although a greater 54 55 proportion of patients withdrew from the classes, stating that they would have preferred 56 57 the web-based programme, it is not known if they would have engaged with and completed 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 the programme. The study showed a lack of engagement in technology in this particular 4 5 population despite a great deal of patient user involvement in the site development. 6 7 8 The trial design meant that patients needed to be willing to be randomised to either group, 9 10 whereas in clinical practice it is more likely that patients will have a preference due to 11 12 genuine choice or practical difficulties that precludes access to supervised rehabilitation 13 14 15 programmes.For This patient peer choice or preference review may improve uptakeonly and completion of a PR 16 17 programme. By exploring alternative forms, such as a web-based programme, patients that 18 19 20 potentially would decline standard PR are provided with an alternative form of intervention. 21 22 Alternative formats of CR, including home-based CR such as The Heart Manual and the 23 24 Angina Plan (41), have been shown to be an effective alternative to conventional CR. In 25 26 27 both randomisation groups, when patients were asked, the highest percent stated their 28 29 preference was the web programme. Those randomised to the conventional PR classes who 30 31 were less disabled (MRC 2) and younger would have preferred the web programme, 32 33 http://bmjopen.bmj.com/ 34 whereas the older patients preferred to attend the classes. Studies have shown that most 35 36 patients with coronary heart disease who are still working prefer to follow a home based 37 38 rehab program instead of conventional supervised classes (42). 39 40 41 Web based rehabilitation may inform the design of future trials. The data collected in this 42 on September 23, 2021 by guest. Protected copyright. 43 study appears to have a role in the delivery of PR. However, a stratified approach may be 44 45 46 needed based upon patient need and choice of delivery to achieve the best outcomes for 47 48 patients and deliver a cost effective model of rehabilitation for a wider population. 49 50 Acknowledgements 51 This article presents independent research funded by the National Institute for Health 52 53 Research (NIHR) under its Research for Patient Benefit (RfPB) Programme (Grant Reference 54 Number PB-PG-0711-25127). The research took place at the University Hospitals of Leicester 55 NHS Trust and was supported by the NIHR Leicester Respiratory Biomedical Research Unit. 56 The views expressed are those of the author(s) and not necessarily those of the NHS, the 57 NIHR or the Department of Health. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 32 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 We thank Aga Glab, Chris Brough and Kelly Edwards for their involvement and support in 4 the study. 5 6 7 Contributors: 8 All authors of the paper have contributed to the design of the work, acquisition, analysis & 9 interpretation of the data. SS, JW, KE, EC, LA, CB, JB and SB were involved in the 10 development of the intervention and design of the trial. EC and SS have been involved in 11 drafting the work or revising it critically for important intellectual content and have given 12 the final approval of the version published. 13 14 15 Funding For peer review only 16 This work was funded by the Research for Patient Benefit (RFPB)(PB-PG-0711-25127) which 17 is part of the funding body National Institute for Health Research (NIHR). The role of the 18 funder did not involve the study design; collection, management, analysis and interpretation 19 20 of data; writing and submission of the report. 21 22 Sponsor 23 The trial is sponsored by University Hospitals of Leicester NHS Trust Research and 24 Development department. 25 26 27 Competing Interests: 28 None 29 30 Ethical approval 31 Ethics approval has been received from the Northampton Research Ethics Committee of the 32 33 UK National Research Ethics Service (Ethics Ref: 12/EM/0351). Written informed consent http://bmjopen.bmj.com/ 34 was obtained from all participants. 35 36 Data sharing statement 37 Additional unpublished data from the study is still being collected and analysed and is only 38 available to members of the study team. 39 40 41 REDCAP 42 Study data were collected and managed using REDCap electronic data capture tools hosted on September 23, 2021 by guest. Protected copyright. 43 at UHL NHS Trust. REDCap (Research Electronic Data Capture) is a secure, web-based 44 application designed to support data capture for research studies, providing 1) an intuitive 45 46 interface for validated data entry; 2) audit trails for tracking data manipulation and export 47 procedures; 3) automated export procedures for seamless data downloads to common 48 statistical packages; and 4) procedures for importing data from external sources. 49 50 51 References 52 53 1.National Institute for Clinical Excellence. Clinical Guideline 101: Chronic obstructive 54 pulmonary disease – Management of chronic obstructive pulmonary disease in adults in 55 primary and secondary care, 21 June 2010. Ref Type: Generic 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 2. Bolton C, Bevan-Smith E, Blakey J, Crave P, Elkin S, Garrod R et al. BTS guidelines on 4 Pulmonary Rehabilitation in Adults. Thorax.(2013)68:ii1–ii30. 5 6 7 3.Keating, A, Lee, A, Holland, A. What prevents people with chronic obstructive pulmonary 8 disease from attending pulmonary rehabilitation? A systematic review. Chronic Respiratory 9 Disease (2011) 8; 2: 89-99. 10 11 4. NHS England: Five Year Forward View. October 2014.Available at 12 https://www.england.nhs.uk/ourwork/futurenhs/ 13 14 15 5. Glasgow, R,For Strycker, L,peer Kurz, D, Faber, review A, Bell, H et al. Recruitment only for an internet-based 16 diabetes self-management program: scientific and ethical implications. 17 Ann.Behav.Med.(2010) 40;1: 40-48. 18 19 20 6. Hoffmann, T, Russell, T, Thompson, L, Vincent, A, Nelson, M. Using the Internet to assess 21 activities of daily living and hand function in people with Parkinson's disease. 22 NeuroRehabilitation.(2008) 23;3 : 253-61. 23 24 7. Allen, M, Iezzoni, L, Huang, A, Huang, L, Leveille, S. Improving patient-clinician 25 communication about chronic conditions: description of an internet-based nurse E-coach 26 27 intervention. Nurs.Res.(2008) 57;2 : 107-12. 28 29 8. Hurkmans, E, van den Berg, M, Ronday, K, Peeters, A, le Cessie, S et al. Maintenance of 30 physical activity after Internet-based physical activity interventions in patients with 31 rheumatoid arthritis. Rheumatology.(2010) 49;1 : 167-72. 32 33 http://bmjopen.bmj.com/ 34 9. Cruz-Correia, R, Fonseca, J, Amaro, M, Lima, L, Araujo, L et al. Web-based or paper-based 35 self-management tools for asthma--patients' opinions and quality of data in a randomized 36 crossover study. Stud.Health Technol.Inform.(2007) 127 :178-89. 37 38 10. Berman, R, Iris, M, Bode, R, Drengenberg, C. The effectiveness of an online mind-body 39 40 intervention for older adults with chronic pain. J.Pain (2009) 10;1 : 68-79. 41 42 11. Walker, E, Wexler,B, DiIorio, C, Escoffery, C, McCarty, F, Yeager, K. Content and on September 23, 2021 by guest. Protected copyright. 43 characteristics of goals created during a self-management intervention for people with 44 epilepsy. J.Neurosci.Nurs.(2009) 41;6 : 312-21. 45 46 47 12. Smeets, T, Brug, J, de Vries, H. Effects of tailoring health messages on physical activity. 48 Health Education Research. (2008) 23;3: 402-413. 49 50 13. Smit, ES, Evers, SM, de Vries, H, Hoving, C. Cost-effectiveness and cost utility of Internet- 51 based computer tailoring for smoking cessation. J Med Internet Res. (2013) 15;3: e57. 52 53 54 14.British Association for Cardiac Prevention and Rehabilitation. Standards and Core 55 components for Cardiovascular Disease Prevention and Rehabilitation; 2012. Available at 56 http://www.bacpr.com 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 32 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 15. Brough C, Boyce S, Houchen-Wolloff L, Sewell L, Singh S. J Evaluating the Interactive 4 Web-Based Program, Activate Your Heart, for Cardiac Rehabilitation Patients: A Pilot Study. 5 6 J. Med Internet Res. (2014)16;10:e242 7 8 16.Devi R, Powell J, Singh S. A Web-Based Program Improves Physical Activity Outcomes in a 9 Primary Care Angina Population: Randomized Controlled Trial. J Med Internet Res (2014) 10 11 16(9):e186. DOI: 10.2196/jmir.3340 12 13 17.Mitchell, K, Johnson-Warrington, V, Apps, L, Bankart, J, Sewell, L, Williams, J et al. A self- 14 management programme for COPD: a randomised controlled trial ERJ (2014) published on 15 line ahead ofFor print doi: 10.1183/09031936.00047814peer review only 16 17 18 18.Apps, L, Mitchell, K, Harrison, S, Sewell, L, Williams, J, Young, H, Steiner, M, Morgan, M, 19 Singh, S.The development and pilot testing of the Self-management Programme of Activity, 20 Coping and Education for Chronic Obstructive Pulmonary Disease (SPACE for COPD). Int. J of 21 COPD (2013) 8; 317-27 22 23 19. Fletcher, C. Standardised questionnaire on respiratory symptoms: a statement prepared 24 25 and approved by the MRC committee on the aetiology of chronic bronchitis (MRC 26 breathlessness score). Br Med J. (1960) 2:1665. 27 28 20. Singh, S, Morgan, M, Scott, S, Walters, D, Hardman, A. Development of a shuttle walking 29 test of disability in patients with chronic airways obstruction. Thorax (1992) 47;12 : 1019-24. 30 31 32 21. Revill, S, Morgan, M, Singh, S, Williams, J, Hardman, A. The endurance shuttle walk: a 33 new field test for the assessment of endurance capacity in chronic obstructive pulmonary http://bmjopen.bmj.com/ 34 disease. Thorax (1999) 54;3 : 213-22. 35 36 22. Burdon J, Juniper E, Kilian K, et al. The perception of breathlessness in asthma. Am Rev 37 38 Respir Dis 1982;126:825-828. 39 40 23.Borg G. Perceived exertion as an indicator of somatic stress. Scand J Rehab Med. 41 1970;2:92-98. on September 23, 2021 by guest. Protected copyright. 42 43 44 24. Williams, J, Singh, S, Sewell, L, Guyatt, G, Morgan, M. Development of a self-reported 45 Chronic Respiratory Questionnaire (CRQ-SR). Thorax (2001) 56;12 : 954-59. 46 47 25. Zigmond, A. S. and R. P. Snaith. The hospital anxiety and depression scale.Acta 48 Psychiatr.Scand. (1983) 67;6 : 361-70. 49 50 51 26. Jones, P, Harding, G, Berry, P, Wiklund, I, Chen, W-H, Kline Leidy, N. Development and 52 first validation of the COPD Assessment Test Eur Respir J (2009) 34:648-654. 53 54 27. Vincent E, Sewell, L., Wagg K, Deacon S, Williams, J., and Singh, S. J. Measuring a change 55 in self efficacy following Pulmonary Rehabilitation: An evaluation of the PRAISE tool. Chest 56 57 (2011) 140;6 1534-1539. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 28. White, R, Walker, P, Roberts, S, Kalisky, S, White, P. Bristol COPD Knowledge 4 Questionnaire (BCKQ): testing what we teach patients about COPD. Chron.Respir.Dis.(2006) 5 3;3 : 123-31. 6 7 8 29. Rutten-van Molken, M, Oostenbrink, J, Tashkin, D, Burkhart, D, Monz, B. Does quality of 9 life of COPD patients as measured by the generic EuroQol five-dimension questionnaire 10 differentiate between COPD severity stages? Chest (2006) 130;4 : 1117-28. 11 12 30. Thompson S and Wordsworth S. An annotated cost questionnaire for completion by 13 14 patients. UK working party on patient costs. 1-3-2001. HERU Discussion Paper 03/01. 15 For peer review only 16 31. Chaplin, E, Hewitt, S, Apps, L, Edwards, K, Brough, C, Glab, A, Bankart, J, Jacobs, R, Boyce, 17 S, Williams, J, Singh, S. The Evaluation of an interactive web-based Pulmonary 18 Rehabilitation programme: protocol for the WEB SPACE for COPD feasibility study. BMJ 19 20 open (2015) 1-7 21 22 32.Steiner M, Holzhauer-Barrie J, Lowe D, Searle L, Skipper E, Welham S, Roberts CM. 23 Pulmonary Rehabilitation: Steps to breathe better. National Chronic Obstructive Pulmonary 24 Disease (COPD) Audit Programme: Clinical audit of Pulmonary Rehabilitation services in 25 England and Wales 2015. London: RCP, 2016. 26 27 www.rcplondon.ac.uk/projects/outputs/pulmonary-rehabilitation-steps-breathe-better 28 29 33.Voncken-Brewster, V, Tange,H, de Vries, H, Nagykaldi, Z, Winkens, B, van der Weijden, T. 30 A randomised controlled trial evaluating the effectiveness of a web-based, computer- 31 tailored self-management intervention for people with or at risk for COPD. Int. J of COPD 32 33 (2015) 10: 1061-1073. http://bmjopen.bmj.com/ 34 35 34 Elfeddali I, Bolman C, Candel MJ, Wiers RW, de Vries H. Preventing smoking relapse via 36 Web-based computer-tailored feedback: a randomized controlled trial. J Med Internet Res. 37 2012;14(4):e109. 38 39 40 35 Strecher VJ, McClure J, Alexander G, et al. The role of engagement in a tailored web- 41 based smoking cessation program: randomized controlled trial. J Med Internet Res. 42 2008;10(5):e36. on September 23, 2021 by guest. Protected copyright. 43 44 36.Pinnock, H, Hanley J, McCloughan L, Todd A, Krishan A, Lewis S et al. Effectiveness of 45 46 telemonitoring integrated into existing clinical services on hospital admission for 47 exacerbation of chronic obstructive pulmonary disease: researcher blind, multicentre, 48 randomised controlled trial. BMJ (2013) 347:1-16. 49 50 37. Singh S, Jones P , Evans R , Morgan M . Minimum clinically important improvement for 51 the incremental shuttle walking test. Thorax (2008) 63 (9): 775-777 52 53 54 38.Murray E. Web-based interventions for behaviour change and self-management: 55 potential, pitfalls, and progress. Med 2.0 (2012)1(2):e3. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 32 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 39.Vandelanotte C, Dwyer T, Van Itallie A, Hanley C, Mummery WK. The development of an 4 internet-based outpatient cardiac rehabilitation intervention: a Delphi study. BMC 5 Cardiovasc Disord (2010)10:27. 6 7 8 40. Gibb, M, Willott, V, Lohar, S, Ward, S, Bolton C , McAlinden, P et al. An evaluation to 9 understand the use of technology within a COPD population. Thorax (2013) 68:A96-A97. 10 11 41.Neville LM, O'Hara B, Milat A. Computer-tailored physical activity behavior change 12 interventions targeting adults: a systematic review. Int J Behav Nutr Phys Act (2009)6:30 13 14 15 42.Grace SL,For McDonald J,peer Fishman D, Caruso review V. Patient preferences only for home-based versus 16 hospital-based cardiac rehabilitation. J Cardiopulm Rehabil (2005) 25;1:24-29. 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 23, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 PR WEB 4 (n = 52) (n = 51) 5 Age (Years) 66.1 ± 8.1 66.4 ± 10.1 6 Gender (% male) 63.5 74.5 7 FEV1 (% predicted) 55.0 ± 20.5 58.7 ± 29.1 8 2 9 BMI (kg/m ) 29.3 ± 6.3 27.9 ± 6.4 10 11 MRC (IQR) 3 (2-4) 3 (2-4) 12 MRC (n) 13 2 21 14 14 3 13 20 15 4 For peer review14 only12 16 5 1 3 17 18 Baseline ISWT (m) 284.2 ± 156.0 296.7 ± 180.8 19 Baseline ESWT (secs) 246.2 ± 144.0 241.7 ± 209.7 20 21 Pre CRQ SR -D 2.7 ± 1.1 2.7 ± 1.2 22 Pre CAT 20.8 ± 7.5 20.8 ± 8.6 23 Pre PRAISE 45.7 ± 7.7 45.6 ± 7.7 24 Pre HADS 25 Anxiety 7.1 ± 5.0 7.9 ± 4.8 26 Depression 5.8 ± 3.6 6.4 ± 3.8 27 Pre BCKQ 37.1 ± 12.5 33.9 ± 8.6 28

29 30 Table 1 Baseline characteristics. Data are presented as n or mean ± SD. FEV1: forced 31 expiratory volume in 1 s; BMI: body mass index; MRC: Medical Research Council; ISWT: 32 Incremental Shuttle Walk Test; ESWT: Endurance Shuttle Walk Test; CRQ-SR- D: Chronic 33 Respiratory disease Questionnaire – Self Report - Dyspnoea; CAT: COPD Assessment Tool; http://bmjopen.bmj.com/ 34 PRAISE: PR Adapted Index of Self Efficacy; HADS: Hospital Anxiety and Depression Scale; 35 36 BCKQ: Bristol COPD Knowledge Questionnaire. 37 38 39 40 41 42 on September 23, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 32 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 Dropouts Completers 4 (n = 29) (n = 22) 5 Gender (M:F) 18:11 20:2 6 7 8 Age (years) 65.3 ± 12 67.6 ± 7 9 10 FEV1 % predicted 63.6 ± 30.2 52.1 ± 27.2 11 12 2 BMI (kg/m ) 29.1 ± 6.7 26.4 ± 5.7 13 14 15 MRC For peer review52% MRC 3 only41% MRC 2 16 27% MRC 3 & 4 17 4% MRC 5 18 Pre ISWT (m) 264.6 334.5 19 20 Pre ESWT (secs) 209.1 278.7 21 22 23 Pre CRQ-D 2.6 2.9 24 25 Pre Anxiety 9.4* 6.5 26 27 28 29 * p < 0.05 between groups 30 31 32 Table 2 Baseline characteristics between WEB completers and dropout 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 23, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 Stage No. of participants 4 No WEB introduction completed 5 5 Not registered 7 6 Stage 1 4 7 Introduction to exercising and goal setting, 8 exercise safety quiz, read educational material 9 10 Stage 2 11 11 Introduction of aerobic exercise programme, set 12 walking target, read educational material 13 Stage 3 2 14 Introduction of strength training programme, 15 set strength target,For continuation peer of aerobic review only 16 training and read education material 17 Stage 4 0 18 Maintain strength and aerobic training, review 19 educational material, knowledge quiz 20 21 22 Table 3 Drop out stages of the WEB programme. 23 24 25 26 27 28 Figure 1 SPACEforCOPD dashboard screen showing tasks completed in stage 1 as well as an 29 overview of exercise progression, goals, knowledge and symptom diary. 30 31 32 Figure 2 Consolidation Standards of Reporting Trials flow diagram of participation. MRC: 33 Medical Research Council; COPD: Chronic Obstructive Pulmonary Disease; PR: Pulmonary http://bmjopen.bmj.com/ 34 Rehabilitation. 35 36 Figure 3 Exercise capacity. Within and Between group changes of the ESWT: endurance 37 shuttle walk test. 38 39 40 Figure 4 Quality of life. Within and Between group changes of CRQ-D: Chronic Respiratory 41 Disease Questionnaire – Self Report – Dyspnoea 42 on September 23, 2021 by guest. Protected copyright. 43 Fig 5a and b Patient preference for Programme Setting Prior to Randomisation 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 32 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 23, 2021 by guest. Protected copyright. 43 44 45 46 47

48 240x389mm (120 x 120 DPI) 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

48 226x290mm (120 x 120 DPI) 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 32 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 23, 2021 by guest. Protected copyright. 42 43 178x184mm (120 x 120 DPI) 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41

on September 23, 2021 by guest. Protected copyright. 42 43 44 45 169x185mm (120 x 120 DPI) 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 32 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 193x115mm (120 x 120 DPI) 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 23, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 32 BMJ Open

1 2 3 4 Supplement for the Web paper BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from 5 6 Stages of the WEBsite 7 8 9 10 SPACEforCOPD® (Figure 1) is an interactive Web-based PR programme that offers a 11 comprehensive package of exercise and education. It was developed following the 12 success of Activate Your Heart which was devised by cardiac rehabilitation specialists 13 14 at our institution, UHL NHS trust. The programme is password protected; each 15 participant was given their own unique password to access the SPACEforCOPD® 16 programme. All participants were able to record and monitor their exercises and 17 strength trainingFor as well peer as interacting review with members ofonly the research team (Figure 18 19 2). 20 21 The SPACEforCOPD® programme was structured to guide the user through four 22 stages that each have specific tasks the user needs to achieve before progressing 23 onto the next stage (Figure 3). Tasks included creating and updating their own short- 24 25 term goals, completing knowledge tests on COPD and exercising safely, and reading 26 specific topics such as inhaler techniques. The educational reading material includes 27 videos and covers topics such as disease education, managing breathlessness, 28 29 recognising the signs and symptoms of an exacerbation, energy conservation and 30 diet and healthy eating (Figure 4). 31 32 In Stage 1, participants were asked to do a multiple choice questionnaire to establish 33 their knowledge regarding the principles of exercising safely. A score of 80% was set 34 35 as a threshold to ensure understanding of these principles. In Stages 2-4, participants 36 were required to record all their exercises, both aerobic and strength, in an exercise 37 diary. During Stage 2, participants were advised to record their aerobic exercise, 5 38 http://bmjopen.bmj.com/ 39 days out of 7. The intensity of the exercise was based on their performance on the 40 baseline ISWT and ESWT exercise tests and prescribed at 85% of baseline 41 performance. In Stage 3, as well as completing 5 out of 7 days of aerobic exercise, 42 patients were asked to also start their resistance training, completing 3 strength 43 44 sessions in a week. Finally, in Stage 4, participants were required to maintain both 45 the aerobic and strength components of the exercise programme for a further 2

46 weeks. There was also interactivity around stress management, knowledge and on September 23, 2021 by guest. Protected copyright. 47 smoking cessation, if appropriate. For smokers, a cost calculator was developed that 48 49 would calculate how much the user had spent or saved since starting the 50 programme. This was delivered along with advice and support to stop smoking. 51 52 Other features embedded within the SPACEforCOPD® programme included a forum 53 54 where patients were also able to share views and experiences with other 55 programme users, a blog, and a frequently asked questions section. The forum was 56 monitored and moderated, as necessary, by the research team. In addition patients 57 were able to communicate privately with the research team via the Ask the Expert 58 59 messaging facility. 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 32

1 2 3 4 The research team members were provided with individual passwords to access the BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from 5 administration section of the programme; this allowed them to view and monitor 6 individual patients’ progress and view patient login data. Patients’ were contacted 7 once a week via email or phone to discuss how they were progressing, any issues or 8 9 barriers to them not continuing the programme. 10 11 All data captured on the programme were encrypted to safeguard patient 12 confidentiality. 13 14 15 16 17 For peer review only 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 http://bmjopen.bmj.com/ 39 40 41 42 43 Figure 1. SPACE for COPD website homepage 44 45 46 on September 23, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 32 BMJ Open

1 2 3 4 BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from 5 6 7 8 9 10 11 12 13 14 15 16 17 For peer review only 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Figure 2. Recording details of exercise and strength training 34 35 36 37 38 http://bmjopen.bmj.com/ 39 40 41 42 43 44 45 46 on September 23, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Figure 3. Part of the dashboard showing what tasks need to be completed in Stage 2

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 32

1 2 3

BMJ Open: first published as 10.1136/bmjopen-2016-013682 on 31 March 2017. Downloaded from 4 5 6 7 8 9 10 11 12 13 14 15 16 17 For peer review only 18 19 20 21 22 23 24 25 26 27 Figure 4. Reading material section 28 29 30 31 32 33 34 35 36 37 38 http://bmjopen.bmj.com/ 39 40 41 42 43 44 45 46 on September 23, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml