HEALTH TECHNOLOGY ASSESSMENT

Transcription

1 HEALTH TECHNOLOGY ASSESSMENT VOLUME 19 ISSUE 75 SEPTEMBER 2015 ISSN Does home oxygen therapy (HOT) in addition to standard care reduce disease severity and improve symptoms in peope with chronic heart faiure? A randomised tria of home oxygen therapy for patients with chronic heart faiure Andrew L Cark, Miriam Johnson, Caroine Fairhurst, David Torgerson, Sarah Cockayne, Sara Rodgers, Susan Griffin, Victoria Agar, Lesey Jones, Samantha Nabb, Ian Harvey, Iain Squire, Jerry Murphy and Michae Greenstone DOI /hta19750

2

3 Does home oxygen therapy (HOT) in addition to standard care reduce disease severity and improve symptoms in peope with chronic heart faiure? A randomised tria of home oxygen therapy for patients with chronic heart faiure Andrew L Cark, 1 * Miriam Johnson, 2 Caroine Fairhurst, 3 David Torgerson, 3 Sarah Cockayne, 3 Sara Rodgers, 3 Susan Griffin, 4 Victoria Agar, 5 Lesey Jones, 6 Samantha Nabb, 7 Ian Harvey, 8 Iain Squire, 9 Jerry Murphy 10 and Michae Greenstone 11 1 Hu York Medica Schoo, Caste Hi Hospita, Cottingham, UK 2 Hu York Medica Schoo, University of Hu, Hu, UK 3 Department of Heath Sciences, York Trias Unit, University of York, York, UK 4 Centre for Heath Economics, University of York, York, UK 5 Hu York Medica Schoo, University of York, York, UK 6 Schoo of Socia Sciences, University of Hu, Hu, UK 7 Department of Sport, Heath and Exercise Science, University of Hu, Hu, UK 8 Department of Academic Cardioogy, Caste Hi Hospita, Cottingham, UK 9 Leicester Cardiovascuar Biomedica Research Unit, Genfied Hospita, Leicester, UK 10 Department of Cardioogy, Darington Memoria Hospita, Darington, UK 11 Medica Chest Unit, Caste Hi Hospita, Cottingham, UK *Corresponding author Decared competing interests of authors: Caroine Fairhurst, Sarah Cockayne, Sara Rodgers and David Torgerson a report other grants from the Nationa Institute for Heath Research Heath Technoogy Assessment programme. Pubished September 2015 DOI: /hta19750

4

5 This report shoud be referenced as foows: Cark AL, Johnson M, Fairhurst C, Torgerson D, Cockayne S, Rodgers S, et a. Does home oxygen therapy (HOT) in addition to standard care reduce disease severity and improve symptoms in peope with chronic heart faiure? A randomised tria of home oxygen therapy for patients with chronic heart faiure. Heath Techno Assess 2015;19(75). Heath Technoogy Assessment is indexed and abstracted in Index Medicus/MEDLINE, Excerpta Medica/EMBASE, Science Citation Index Expanded (SciSearch ) and Current Contents / Cinica Medicine.

6

7 Heath Technoogy Assessment HTA/HTA TAR ISSN (Print) ISSN (Onine) Impact factor: Heath Technoogy Assessment is indexed in MEDLINE, CINAHL, EMBASE, The Cochrane Library and the ISI Science Citation Index. This journa is a member of and subscribes to the principes of the Committee on Pubication Ethics (COPE) ( Editoria contact: [email protected] The fu HTA archive is freey avaiabe to view onine at Print-on-demand copies can be purchased from the report pages of the NIHR Journas Library website: Criteria for incusion in the Heath Technoogy Assessment journa Reports are pubished in Heath Technoogy Assessment (HTA) if (1) they have resuted from work for the HTA programme, and (2) they are of a sufficienty high scientific quaity as assessed by the reviewers and editors. Reviews in Heath Technoogy Assessment are termed systematic when the account of the search appraisa and synthesis methods (to minimise biases and random errors) woud, in theory, permit the repication of the review by others. HTA programme The HTA programme, part of the Nationa Institute for Heath Research (NIHR), was set up in It produces high-quaity research information on the effectiveness, costs and broader impact of heath technoogies for those who use, manage and provide care in the NHS. Heath technoogies are broady defined as a interventions used to promote heath, prevent and treat disease, and improve rehabiitation and ong-term care. The journa is indexed in NHS Evidence via its abstracts incuded in MEDLINE and its Technoogy Assessment Reports inform Nationa Institute for Heath and Care Exceence (NICE) guidance. HTA research is aso an important source of evidence for Nationa Screening Committee (NSC) poicy decisions. For more information about the HTA programme pease visit the website: This report The research reported in this issue of the journa was funded by the HTA programme as project number 06/80/01. The contractua start date was in January The draft report began editoria review in November 2014 and was accepted for pubication in May The authors have been whoy responsibe for a data coection, anaysis and interpretation, and for writing up their work. The HTA editors and pubisher have tried to ensure the accuracy of the authors report and woud ike to thank the reviewers for their constructive comments on the draft document. However, they do not accept iabiity for damages or osses arising from materia pubished in this report. This report presents independent research funded by the Nationa Institute for Heath Research (NIHR). The views and opinions expressed by authors in this pubication are those of the authors and do not necessariy refect those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Heath. If there are verbatim quotations incuded in this pubication the views and opinions expressed by the interviewees are those of the interviewees and do not necessariy refect those of the authors, those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Heath. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. Pubished by the NIHR Journas Library ( produced by Prepress Projects Ltd, Perth, Scotand (

8 Editor-in-Chief of Heath Technoogy Assessment and NIHR Journas Library Professor Tom Waey Director, NIHR Evauation, Trias and Studies and Director of the HTA Programme, UK NIHR Journas Library Editors Professor Ken Stein Chair of HTA Editoria Board and Professor of Pubic Heath, University of Exeter Medica Schoo, UK Professor Andree Le May Chair of NIHR Journas Library Editoria Group (EME, HS&DR, PGfAR, PHR journas) Dr Martin Ashton-Key Consutant in Pubic Heath Medicine/Consutant Advisor, NETSCC, UK Professor Matthias Beck Chair in Pubic Sector Management and Subject Leader (Management Group), Queen s University Management Schoo, Queen s University Befast, UK Professor Aieen Carke Professor of Pubic Heath and Heath Services Research, Warwick Medica Schoo, University of Warwick, UK Dr Tessa Criy Director, Crysta Bue Consuting Ltd, UK Dr Peter Davidson Director of NETSCC, HTA, UK Ms Tara Lamont Scientific Advisor, NETSCC, UK Professor Eaine McCo Director, Newcaste Cinica Trias Unit, Institute of Heath and Society, Newcaste University, UK Professor Wiiam McGuire Professor of Chid Heath, Hu York Medica Schoo, University of York, UK Professor Geoffrey Meads Professor of Heath Sciences Research, Facuty of Education, University of Winchester, UK Professor John Norrie Heath Services Research Unit, University of Aberdeen, UK Professor John Powe Consutant Cinica Adviser, Nationa Institute for Heath and Care Exceence (NICE), UK Professor James Raftery Professor of Heath Technoogy Assessment, Wessex Institute, Facuty of Medicine, University of Southampton, UK Dr Rob Riemsma Reviews Manager, Keijnen Systematic Reviews Ltd, UK Professor Heen Roberts Professor of Chid Heath Research, UCL Institute of Chid Heath, UK Professor Heen Snooks Professor of Heath Services Research, Institute of Life Science, Coege of Medicine, Swansea University, UK Professor Jim Thornton Professor of Obstetrics and Gynaecoogy, Facuty of Medicine and Heath Sciences, University of Nottingham, UK Pease visit the website for a ist of members of the NIHR Journas Library Board: Editoria contact: [email protected] NIHR Journas Library

9 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Abstract Does home oxygen therapy (HOT) in addition to standard care reduce disease severity and improve symptoms in peope with chronic heart faiure? A randomised tria of home oxygen therapy for patients with chronic heart faiure Andrew L Cark, 1* Miriam Johnson, 2 Caroine Fairhurst, 3 David Torgerson, 3 Sarah Cockayne, 3 Sara Rodgers, 3 Susan Griffin, 4 Victoria Agar, 5 Lesey Jones, 6 Samantha Nabb, 7 Ian Harvey, 8 Iain Squire, 9 Jerry Murphy 10 and Michae Greenstone 11 1 Hu York Medica Schoo, Caste Hi Hospita, Cottingham, UK 2 Hu York Medica Schoo, University of Hu, Hu, UK 3 Department of Heath Sciences, York Trias Unit, University of York, York, UK 4 Centre for Heath Economics, University of York, York, UK 5 Hu York Medica Schoo, University of York, York, UK 6 Schoo of Socia Sciences, University of Hu, Hu, UK 7 Department of Sport, Heath and Exercise Science, University of Hu, Hu, UK 8 Department of Academic Cardioogy, Caste Hi Hospita, Cottingham, UK 9 Leicester Cardiovascuar Biomedica Research Unit, Genfied Hospita, Leicester, UK 10 Department of Cardioogy, Darington Memoria Hospita, Darington, UK 11 Medica Chest Unit, Caste Hi Hospita, Cottingham, UK *Corresponding author [email protected] Background: Home oxygen therapy (HOT) is commony used for patients with severe chronic heart faiure (CHF) who have intractabe breathessness. There is no tria evidence to support its use. Objectives: To detect whether or not there was a quaity-of-ife benefit from HOT given as ong-term oxygen therapy (LTOT) for at east 15 hours per day in the home, incuding overnight hours, compared with best medica therapy (BMT) in patients with severey symptomatic CHF. Design: A pragmatic, two-arm, randomised controed tria recruiting patients with severe CHF. It incuded a inked quaitative substudy to assess the views of patients using home oxygen, and a free-standing substudy to assess the haemodynamic effects of acute oxygen administration. Setting: Heart faiure outpatient cinics in hospita or the community, in a range of urban and rura settings. Participants: Patients had to have heart faiure from any aetioogy, New York Heart Association (NYHA) cass III/IV symptoms, at east moderate eft ventricuar systoic dysfunction, and be receiving maximay toerated medica management. Patients were excuded if they had had a cardiac resynchronisation therapy device impanted within the past 3 months, chronic obstructive pumonary disease fufiing the criteria for LTOT or maignant disease that woud impair surviva or were using a device or medication that woud impede their abiity to use LTOT. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. vii

10 ABSTRACT Interventions: Patients received BMT and were randomised (unbinded) to open-abe LTOT, prescribed for 15 hours per day incuding overnight hours, or no oxygen therapy. Main outcome measures: The primary end point was quaity of ife as measured by the Minnesota Living with Heart Faiure (MLwHF) questionnaire score at 6 months. Secondary outcomes incuded assessing the effect of LTOT on patient symptoms and disease severity, and assessing its acceptabiity to patients and carers. Resuts: Between Apri 2012 and February 2014, 114 patients were randomised to receive either LTOT or BMT. The mean age was 72.3 years [standard deviation (SD) 11.3 years] and 70% were mae. Ischaemic heart disease was the cause of heart faiure in 84%; 95% were in NYHA cass III; the mean eft ventricuar ejection fraction was 27.8%; and the median N-termina pro-b-type natriuretic hormone was 2203 ng/. The primary anaysis used a covariance pattern mixed mode which incuded patients ony if they provided data for a baseine covariates adjusted for in the mode and outcome data for at east one post-randomisation time point (n = 102: intervention, n = 51; contro, n = 51). There was no difference in the MLwHF questionnaire score at 6 months between the two arms [at baseine the mean score was 54.0 (SD 18.4) for LTOT and 54.0 (SD 17.9) for BMT; at 6 months the mean score was 48.1 (SD 18.5) for LTOT and 49.0 (SD 20.2) for BMT; adjusted mean difference 0.10, 95% confidence interva (CI) 6.88 to 6.69; p = 0.98]. At 3 months, the adjusted mean MLwHF questionnaire score was ower in the LTOT group ( 5.47, 95% CI to 0.41; p = 0.03) and breathessness scores improved, athough the effect did not persist to 6 months. There was no effect of LTOT on any secondary measure. There was a greater number of deaths in the BMT arm (n = 12 vs. n = 6). Adherence was poor, with ony 11% of patients reporting using the oxygen as prescribed. Concusions: Athough the study was significanty underpowered, HOT prescribed for 15 hours per day and subsequenty used for a mean of 5.4 hours per day has no impact on quaity of ife as measured by the MLwHF questionnaire score at 6 months. Suggestions for future research incude (1) a tria of patients with severe heart faiure randomised to have emergency oxygen suppy in the house, suppied by cyinders rather than an oxygen concentrator, powered to detect a reduction in admissions to hospita, and (2) a study of bed-bound patients with heart faiure who are in the ast few weeks of ife, powered to detect changes in symptom severity. Tria registration: Current Controed Trias ISRCTN Funding: This project was funded by the NIHR Heath Technoogy Assessment programme and wi be pubished in fu in Heath Technoogy Assessment; Vo. 19, No. 75. See the NIHR Journas Library website for further project information. viii NIHR Journas Library

11 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Contents List of tabes List of figures List of abbreviations Pain Engish summary Scientific summary xiii xv xvii xix xxi Chapter 1 Introduction 1 Heart faiure 1 Pumonary disease and oxygen 2 Heart disease and oxygen 3 Seep apnoea 3 Home oxygen therapy for breathessness 4 The home oxygen therapy tria rationae 5 Chapter 2 Synopsis of tria evoution 7 Tria structure and protoco 7 Stage 1: the North East Oxygen Network tria 7 Stage 2: the three-arm home oxygen therapy tria 9 Stage 3: the two-arm home oxygen therapy tria 10 Chapter 3 The home oxygen therapy tria: methods 11 Approvas obtained 11 Patient study group 12 Incusion criteria 12 Excusion criteria 12 Randomisation 12 Primary outcome 12 Power cacuation 13 Secondary outcomes 13 Measures of quaity of ife 13 Severity of heart faiure 14 Exercise capacity 14 Performance 14 Comorbidity 14 Prevaence of hypoxia 14 Safety and adherence 15 Other measurements 15 Cost-effectiveness 15 Adverse events 16 Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. ix

12 CONTENTS Chapter 4 Statistica anaysis 17 Tria competion 17 Baseine data 17 Primary anaysis 17 Sensitivity anayses 18 Secondary anaysis 18 Comparisons with the nocturna oxygen therapy subgroup 18 Secondary outcomes 18 Mortaity 18 Number of days aive and out of hospita 19 Heath-reated quaity of ife as measured by the European Quaity of Life-5 Dimensions 19 Prevaence of hypoxia 19 Patient-reported adherence 19 Number of hours of oxygen used measured by concentrator meter 19 Adverse events 19 Chapter 5 Study procedures 21 Baseine and foow-up assessments 21 Home oxygen therapy 21 Best medica therapy 22 Foow-up 22 Tria competion 23 Chapter 6 Resuts 25 Tria recruitment 25 Withdrawas 28 Baseine participant characteristics 28 Primary outcome 31 Predictors of Minnesota Living with Heart Faiure questionnaire score at 6 months 34 Minnesota Living with Heart Faiure questionnaire score at 3 and 12 months 35 Minnesota Living with Heart Faiure questionnaire score at 6 months adjusting for centre 35 The nocturna oxygen therapy subgroup 35 Secondary outcomes 38 Epworth Seepiness Scae 38 Numerica Rating Scae for breathessness 38 Hospita Anxiety and Depression Scae 40 Karnofsky Performance Status scae of physica activity 41 Charson Comorbidity Index 41 The 6-minute wak test 41 Prevaence of hypoxia 44 N-termina pro-b-type natriuretic hormone 44 Left ventricuar ejection fraction 47 Mortaity 47 Number of days aive and out of hospita 50 European Quaity of Life-5 Dimensions 50 Adverse events 51 Serious adverse events 51 Foow-up for serious adverse events 53 Non-serious adverse events 53 Adherence 54 Patient-reported adherence 55 Oxygen suppiers meter readings 56 Overnight Embetta seep study 57 x NIHR Journas Library

13 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Chapter 7 Acute oxygen substudy 61 Background 61 Methods 61 Incusion criteria 61 Excusion criteria 62 Procedures 62 Sampe size 62 Cacuations 62 Resuts 63 Chapter 8 The home oxygen therapy tria quaitative substudy 67 Background and rationae 67 Aims and objectives 67 Methods 67 Design 67 Samping strategy 68 Data coection 68 Data anaysis 68 Resuts 69 Participants 69 Findings 70 Chapter 9 Discussion 81 Cinica effectiveness 81 Quaity of ife measures 82 Safety 82 Why was the home oxygen therapy tria neutra? 82 Patient seection 83 Prevaence of hypoxia 83 Adherence 83 Substudies 85 Acute oxygen substudy (see Chapter 7) 85 Quaitative substudy (see Chapter 8) 85 Limitations 86 Concusions 88 Recommendations for future research 88 Acknowedgements 89 References 93 Appendix 1 Origina Heath Technoogy Assessment commissioning brief 101 Appendix 2 Reguatory approvas and detais of study sites 105 Appendix 3 Home oxygen therapy patient information sheet version Appendix 4 Patient consent form 117 \ Appendix 5 Graphica checks of the assumptions for the primary anaysis mode 119 Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xi

14

15 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 List of tabes TABLE 1 The NYHA cassification of symptoms in heart faiure 1 TABLE 2 Schedue of patient assessments: study structure 22 TABLE 3 Number of participants randomised by group and site, n (%) 26 TABLE 4 Reasons (where given) for participant change of circumstances 28 TABLE 5 Baseine participant characteristics by treatment group 29 TABLE 6 Baseine characteristics by treatment group 30 TABLE 7 Raw and adjusted primary outcome data 32 TABLE 8 Baseine characteristics of participants as incuded in the primary anaysis by treatment group 33 TABLE 9 Coefficients from the primary anaysis mode 34 TABLE 10 Baseine participant characteristics of a patients randomised up to 30 Apri 2013 by treatment group 36 TABLE 11 Observed and adjusted summaries for the primary outcome measure by treatment group at each time point 37 TABLE 12 Observed summaries of the ESS by treatment group at each time point 38 TABLE 13 The NRS for breathessness score by treatment group at each time point 38 TABLE 14 Observed summaries of the HADS anxiety subscae by treatment group at each time point 40 TABLE 15 Observed summaries of the HADS depression subscae by treatment group at each time point 41 TABLE 16 Distribution of KPS scores at baseine 42 TABLE 17 Distribution of KPS scores at baseine for patients recruited by 1 May TABLE 18 Karnofsky Performance Status scae summaries by treatment group at each time point 43 TABLE 19 Charson Comorbidity Index scores by treatment group at each time point 44 TABLE 20 Six-minute wak test distance by treatment group at each time point 44 TABLE 21 Arteria oxygen saturation in the 6MWT 45 Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xiii

16 LIST OF TABLES TABLE 22 Oxygen saturation measured before the 6MWT, by treatment group at each time point 45 TABLE 23 Oxygen saturation measured at peak (maximum saturation observed during 6MWT), by treatment group at each time point 46 TABLE 24 Oxygen saturation measured at 5 minutes post 6MWT, by treatment group at each time point 46 TABLE 25 Leve of NT-proBNP by treatment group at each time point 46 TABLE 26 Left ventricuar ejection fraction and severity of LV dysfunction by treatment group at each time point 48 TABLE 27 Days aive and out of hospita 50 TABLE 28 Summary of EQ-5D-3L scores for each treatment group by time point 51 TABLE 29 Serious adverse events 52 TABLE 30 Foow-up SAEs 53 TABLE 31 Non-SAEs 54 TABLE 32 Patient-reported adherence to HOT 55 TABLE 33 Oxygen machine usage data 56 TABLE 34 Resuts from the Embetta seep test by treatment group at each time point 58 TABLE 35 Baseine data for patients in the oxygen substudy 63 TABLE 36 Pressure measurements during right heart catheterisation 64 TABLE 37 Oxygen saturation and derived measures from cardiac catheterisation 64 TABLE 38 Patients in the quaitative substudy 69 TABLE 39 Major themes and subthemes 70 TABLE 40 Studies of adherence in patients receiving LTOT for chronic airways disease 84 TABLE 41 Timing of MREC approvas 105 TABLE 42 Timings of approvas for sites 105 xiv NIHR Journas Library

17 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 List of figures FIGURE 1 The origina design for the NEON tria 7 FIGURE 2 The origina design for the three-arm HOT tria 9 FIGURE 3 The fina design for the two-arm HOT tria 10 FIGURE 4 The overa rate of recruitment in the HOT tria 25 FIGURE 5 Participant recruitment by site (two sites did not recruit any patients) 26 FIGURE 6 The CONSORT fow diagram 27 FIGURE 7 Adjusted means for MLwHF questionnaire score by treatment group over time from the primary anaysis mode 34 FIGURE 8 Kapan Meier surviva curve 49 FIGURE 9 Kapan Meier surviva curve for BMT vs. LTOT pus NOT 49 FIGURE 10 Reations between haemodynamic variabes 65 FIGURE 11 Histogram of the standardised residuas 119 FIGURE 12 Q Q pot of the standardised residuas 120 FIGURE 13 Scatterpot of residuas vs. predicted vaues 120 Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xv

18

19 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 List of abbreviations 6MWT 6-minute wak test HRQoL heath-reated quaity of ife AHI apnoea hypopnoea index HTA Heath Technoogy Assessment AMD adjusted mean difference KPS Karnofsky Performance Status scae ANCOVA anaysis of covariance LTOT ong-term oxygen therapy BMT best medica therapy LV eft ventricuar CCI Charson Comorbidity Index LVEF eft ventricuar ejection fraction CHF chronic heart faiure MLwHF Minnesota Living with Heart Faiure CI confidence interva MRC Medica Research Counci CONSORT Consoidated Standards of Reporting Trias MREC Medica Research and Ethics Committee COPD DAOH df ECG chronic obstructive pumonary disease days aive and out of hospita degree of freedom eectrocardiography NEON NOT NRS NT-proBNP North East Oxygen Network nocturna oxygen therapy Numerica Rating Scae N-termina pro-b-type natriuretic hormone EQ-5D EQ-5D-3L ESS FEV 1 European Quaity of Life-5 Dimensions European Quaity of Life-5 Dimensions-3 Leves Epworth Seepiness Scae forced expiratory voume in 1 second NYHA PaO 2 QoL RCT SAE SD New York Heart Association partia pressure of arteria oxygen quaity of ife randomised controed tria serious adverse event standard deviation FVC forced vita capacity SDB seep-disordered breathing HADS HOT Hospita Anxiety and Depression Scae home oxygen therapy SE SVR standard error systemic vascuar resistance HR hazard ratio Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xvii

20

21 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Pain Engish summary Oxygen therapy is the administration of additiona oxygen for medica reasons. Patients with severe chronic heart faiure suffer from breathessness that may ruin their quaity of ife (QoL). Party because patients with severe ung disease benefit from home oxygen therapy (HOT), patients with severe heart faiure are often prescribed home oxygen. However, oxygen therapy can be burdensome. It imits mobiity, it can cause soreness around the nose and the equipment is noisy. There is no evidence to support its use in patients with heart faiure. The HOT tria was designed to measure any beneficia effects on QoL measured with the Minnesota Living with Heart Faiure questionnaire. We aocated, at random, 114 patients with severey symptomatic heart faiure either to receive home oxygen for 15 hours a day or not to receive oxygen therapy. A participants continued to receive the best medica therapy for their condition. The average age of patients was 70 years, and 70% of patients were men. None of the patients had a ow eve of oxygen in their bood. As ony 11% of patients reported that they used the oxygen for the fu 15 hours a day, the tria was stopped eary. We found no evidence that home oxygen improved patients QoL, symptoms or any other measurement of severity of heart faiure. There was a sma improvement in surviva with oxygen, but the difference was not statisticay significant. Further study might identify whether or not having emergency oxygen avaiabe at home woud reduce the need for admission to hospita. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xix

22

23 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Scientific summary Background Chronic heart faiure (CHF) affects at east 1% of the popuation and is responsibe for around 4% of a admissions to hospita in the UK. The prognosis of heart faiure if it is not we treated is beak. The cinica course for most patients with heart faiure tends to be one of gradua decine, often punctuated with episodes of severe deterioration resuting in hospitaisation. Towards the end of their ives, many patients with CHF become very symptomatic, with imiting breathessness on minima exertion and even at rest. Athough standard treatment may reieve symptoms, for many the ast few months and even years of ife can be miserabe, with persisting severe breathessness on minima exertion and episodic hospitaisation. Home oxygen therapy (HOT) is commony prescribed to patients with severey symptomatic CHF in order to aeviate suffering. However, unike the situation for patients with chronic obstructive airways disease and severe hypoxia, in whom oxygen proongs surviva, there is no evidence to support the use of HOT in patients with CHF. Objectives The HOT tria was designed to address the question of whether or not there is any effect of HOT on quaity of ife (QoL) in patients with severey symptomatic heart faiure. Secondary end points were to assess the effects of HOT on breathessness, 6-minute wak test distance, severity of eft ventricuar (LV) systoic dysfunction, N-termina B-type natriuretic hormone (NT-proBNP) eve and prognosis. The study consisted of three parts. The main study was a randomised controed tria (RCT) designed to measure the impact of HOT on QoL in severey symptomatic patients. A quaitative substudy assessed the burden on patients and their carers, and an acute oxygen substudy assessed whether or not there was any effect of oxygen given in the same concentration as used by concentrators at home on haemodynamics. Methods The main study was a pragmatic, two-arm RCT, recruiting patients with severe CHF. Patients were recruited from heart faiure outpatient cinics in hospita or the community, in a range of urban and rura settings. Patients had to have heart faiure of any aetioogy, severe symptoms (breathessness either at rest or on minima exertion) and at east moderate LV systoic dysfunction, and be receiving maximay toerated medica management. Patients were excuded if they had had a cardiac resynchronisation therapy device impanted in the past 3 months, chronic obstructive pumonary disease fufiing the criteria for ong-term oxygen therapy (LTOT), interstitia ung disease or maignant disease that woud impair surviva or were using a device or medication that woud impede their abiity to use LTOT. Patients received best medica therapy (BMT) and were randomised to open-abe LTOT, prescribed for 15 hours per day incuding overnight hours, or no oxygen therapy. Home oxygen was deivered by concentrators in the patients homes. The inspired oxygen was increased from 20.9% (norma room air) to approximatey 28%. There were two substudies: a inked quaitative substudy to assess the view of 25 patients and a free-standing oxygen substudy to assess the haemodynamic effects of acute oxygen administration. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xxi

24 SCIENTIFIC SUMMARY Resuts The HOT tria was stopped eary by the funders, the Heath Technoogy Assessment programme, because of poor patient adherence to the oxygen prescription. Between Apri 2012 and February 2014, 114 patients were randomised to receive either LTOT or BMT. The mean age was 72.3 years [standard deviation (SD) 11.3 years] and 70% of patients were mae. Ischaemic heart disease was the cause of heart faiure in 84% of patients; 95% were in New York Heart Association cass III; mean eft ventricuar ejection fraction (LVEF) was 27.8%; and median NT-proBNP was 2203 ng/. Arteria oxygen saturation was norma at rest and there was no significant change in arteria oxygen saturation during exercise or during recovery from exercise. There was a ow prevaence of seep-disordered breathing. The primary anaysis used a covariance pattern mixed mode which incuded patients ony if they provided data for a baseine covariates adjusted for in the mode and outcome data for at east one post-randomisation time point (n = 102: intervention, n = 51; contro, n = 51). There was no difference in Minnesota Living with Heart Faiure (MLwHF) questionnaire score at 6 months between the two arms [at baseine the mean score was 54.0 (SD 18.4) for LTOT and 54.0 (SD 17.9) for BMT; at 6 months the mean score was 48.1 (SD 18.5) for LTOT and 49.0 (SD 20.2) for BMT; adjusted mean difference 0.10, 95% confidence interva (CI) 6.88 to 6.69; p = 0.98]. At 3 months, the adjusted mean MLwHF questionnaire score was ower in the LTOT group (adjusted mean difference 5.47, 95% CI to 0.41; p = 0.03), coinciding with improvements in breathessness scores which did not persist to 6 months. There was no effect of LTOT on any secondary measure incuding 6-minute wak test distance, NT-proBNP eve and LVEF. There was sighty greater surviva in the oxygen-treated group (unadjusted hazard ratio 2.03, 95% CI 0.76 to 5.40, for BMT reative to LTOT), but the difference was not statisticay significant (p = 0.16). In the haemodynamic substudy there were no deeterious effects of 28% oxygen. There was a sma increase in cardiac output and a sma fa in pumonary vascuar resistance. Adherence to HOT was poor, with ony 11% of patients reporting using the oxygen as prescribed. Findings from the quaitative substudy suggested that participants viewed study participation in the tria both as an atruistic act and as a way of accessing optima cinica care. Adherence was reated not specificay to the context of a cinica tria but to a deep-seated beief that oxygen was a therapy for acute deterioration or for those with end-stage disease. Thus, participants fet that the use of LTOT was counterintuitive, despite cear expanation of the tria s aim. This misunderstanding formed a poor basis for subsequent weighing up by the participants of the benefit burden baance of the LTOT. Concusions The prevaence of hypoxia in patients with severe heart faiure at rest, foowing exercise and during an overnight seep test is ow. There is no evidence that LTOT, athough safe, improves the symptoms, prognosis or severity of heart faiure in patients with severe CHF at 6 months. There is no evidence to support the use of HOT in patients with heart faiure. xxii NIHR Journas Library

25 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Recommendations for future research The tria was stopped eary because of poor adherence to the prescription of 15 hours per day. However, the prescription was based on extrapoation from studies of patients with a different pathoogy, chronic airways disease, and who had severe hypoxia. It may be that shorter periods of exposure might have been effective, either in terms of symptom reief or in terms of preventing hospitaisation. We suggest that two further studies might be appropriate: 1. a tria of patients with severe heart faiure randomised to have emergency oxygen suppy in the house, suppied by cyinders rather than oxygen concentrator, powered to detect a reduction in admissions to hospita 2. a study of bed-bound patients with heart faiure who are in the ast few weeks of ife, powered to detect changes in symptom severity. Tria registration This tria is registered as ISRCTN Funding Funding for this study was provided by the Heath Technoogy Assessment programme of the Nationa Institute for Heath Research. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. xxiii

26

27 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Chapter 1 Introduction Heart faiure Heart faiure is a cinica syndrome that arises when the heart fais to pump in a manner adequate to meet the body s needs. It is increasingy common and affects between 1% and 2% of the UK popuation. Its incidence and prevaence rise markedy with age. 1 The most common cause of heart faiure is myocardia infarction and, as more peope survive acute myocardia infarction with modern therapy, so the popuation of patients with damaged heart musces grows. 2 Heart faiure is the singe most common cause for admission to hospita in Engand and Waes and, foowing admission to hospita, there is a 25% chance of readmission or death within 12 weeks. By 1 month after an index admission, 15% of patients have died, either as an inpatient or during the days foowing discharge. 3 The prognosis of heart faiure is beak if it is not we treated. However, one of the greatest success stories of modern medicine is the dramatic improvement in prognosis for patients with chronic heart faiure (CHF). Good medica management argey consists of medicines designed to bock the adverse consequences of neuroendocrine activation such as beta-bockers, angiotensin-converting enzyme inhibitors and mineraocorticoid receptor antagonists. In seected patients, treatment with cardiac resynchronisation therapy aso improves prognosis and, taken together, these treatments approximatey doube ife expectancy. 4 Chronic heart faiure has been recognised for many years as having the greatest symptomatic burden of any chronic medica condition. 5 The cardina symptoms of heart faiure are breathessness and fatigue, particuary on exertion. Worsening breathessness is part of the cause of most admissions to hospita with heart faiure, athough many patients aso compain of anke sweing due to fuid retention. Drug therapy is very successfu in controing symptoms, and can induce remission in a number of patients; that is, their symptoms can remit amost entirey. However, the cinica course for most patients with heart faiure tends to be one of gradua decine, often punctuated with episodes of severe deterioration resuting in hospitaisation. Symptom severity is most commony measured using the New York Heart Association (NYHA) cassification of symptoms (Tabe 1). Towards the end of their ives, many patients with CHF become very symptomatic, with imiting breathessness on minima exertion (cass III) and even at rest (cass IV). Once they have reached this stage, athough patients need continued treatment with drugs known to improve prognosis, the emphasis of treatment becomes the reief of symptoms, that is, paiative care. However, athough drug treatment with diuretics (which reieve fuid congestion), other drugs and pacing devices may reieve symptoms, for many the ast few months and even years of ife can be miserabe, with persisting severe breathessness on minima exertion and episodic hospitaisation. TABLE 1 The NYHA cassification of symptoms in heart faiure NYHA cass I II III IV Symptoms Breathess on severe exertion (norma) Breathess and/or fatigue on moderate exertion Breathess and/or fatigue on mid exertion Breathess and/or fatigue at rest Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 1

28 INTRODUCTION Athough there is some evidence that opioids may reieve breathessness in patients with chronic airways disease and cancer, 6 8 the evidence is mixed in heart faiure, 9 11 and there is no specific intervention for the intractabe breathessness of severe CHF. Another frequenty encountered group with severe breathessness is those patients with chronic airways disease. In patients with chronic airways disease who aso have hypoxia, there is reasonaby robust evidence that ong-term oxygen therapy (LTOT) can improve prognosis as we as symptoms (see Pumonary disease and oxygen). By extension from these data, physicians often use home oxygen therapy (HOT) for patients with severey symptomatic heart faiure. There is, however, no evidence that LTOT is hepfu in CHF, either for the reief of symptoms or to improve prognosis. Pumonary disease and oxygen For many years, patients with chronic airways disease or chronic obstructive pumonary disease (COPD) have been treated with LTOT, particuary if they have evidence of hypoxia at rest. Treatment is given for at east 15 hours a day (incuding overnight). The evidence for the benefit of oxygen therapy comes from randomised cinica trias; the Medica Research Counci s (MRC s) oxygen tria 12 and the Nocturna Oxygen Therapy (NOT) tria 13 are perhaps the best known. A Cochrane review of oxygen therapy for patients with chronic airways disease suggests that ong[-]term home oxygen therapy improved surviva in... COPD patients with severe hypoxaemia (PaO 2 [partia pressure of arteria oxygen] ess than 55 mmhg (8.0 kpa)). 14 In the MRC s oxygen tria, 12 treating five patients with severe hypoxaemic COPD with LTOT saved one ife over the 5-year study period. 14 The prognostic benefits were not apparent unti after more than 1 year of therapy. Athough it does not affect prognosis in peope with more modest hypoxaemia, LTOT does appear to hep by reducing the severity of breathessness. 15 There is no evidence that NOT aone (in other words, giving oxygen ony at night) improves prognosis. 14 The authors of the systematic review and meta-anaysis observed significant heterogeneity in most of their anayses and pointed out that most studies were either singe binded or not binded at a. They therefore recommended an individua approach to care unti data from arge randomised controed trias (RCTs) are avaiabe. Data on the effects of oxygen therapy on quaity of ife (QoL) in patients with chronic airways disease are not cear-cut. Athough the MRC reported that symptoms improved, few data were given. In patients with moderate hypoxaemia, the Cochrane meta-anaysis reported a reasonaby robust improvement in breathessness equivaent to a reduction of 0.78 cm on a 10-cm visua anaogue scae. 15 It is difficut to estimate adherence to LTOT in peope with airways disease. Using the oxygen deivery system for 15 hours per day is ceary burdensome, and most studies suggest that adherence to this demand is ess than 50%. 16 The summary figure of 45 70% is commony quoted, 17 but the studies from which the figures are derived are now quite od (see Chapter 9 for discussion). The ony recent study suggests that adherence remains poor. 18 Athough the origina studies demonstrated a positive reationship between benefit and duration of oxygen, the mechanism is not cear. Peope who used oxygen for a onger period of time woud have been more ikey to have prevented worse desaturations during seep or exertion, and the same benefit coud have been achieved by suppementa oxygen at night ony, or during exertion. Furthermore, if the prevention of exertion-induced desaturation heped exercise toerance, then increased physica activity and reconditioning over time coud have been the mechanism of improved symptoms and prognosis. 19 However, given that the ony studies to show an improvement in prognosis had a target of oxygen use for ong periods of time during the 24 hours, 15 hours per day remains the recommended prescription for prognostic benefit. 2 NIHR Journas Library

29 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Heart disease and oxygen Oxygen is commony prescribed for patients with heart disease. There is a widespread perception that oxygen therapy can do no harm and may possiby be hepfu and, thus, patients are often given high concentrations of inspired oxygen immediatey foowing acute myocardia infarction or when they are admitted with acute pumonary oedema. It is aso commony used during an admission for CHF. Patients with severe (or even end-stage) CHF can appear very simiar to patients with severe chronic airways disease. They are breathess at rest or on minima exertion despite maxima medica therapy. A consequence is that HOT is often prescribed for severey breathess patients with heart faiure, even in the absence of hypoxia, particuary towards the end of ife. There is ony very imited evidence for the use of oxygen in heart disease, and much of the evidence suggests that oxygen might be harmfu. In a study of patients with acute myocardia infarction, oxygen was given at as near to 100% as possibe. Oxygen therapy was associated with a fa in cardiac output and stroke voume, together with a rise in heart rate. 20 The faiing heart requires a higher fiing pressure. The higher the fiing pressure, the worse the cardiac function; hence an intervention causing a rise in fiing pressure is deeterious. In a study of 10 patients with CHF, % oxygen caused a rise in cardiac fiing pressure, a fa in cardiac output and an increase in systemic vascuar resistance (SVR). The SVR represents the oad against which the heart has to work: the higher the SVR, the greater the work required of the heart. In another study of 12 patients with CHF, 22 high-dose oxygen was associated with an increase in eft ventricuar (LV) fiing pressure. In contrast, in a study of patients with CHF given ower doses of oxygen (50%), exercise capacity increased and patients were ess breathess and had a ower eve of ventiation during exercise than during exercise with room air. 23 Findings from studies are inconsistent; another study has suggested that suppementary oxygen has itte effect on exercise performance. 24 There is itte evidence of the effect of oxygen when given to patients with heart faiure at the much more modest eves used for treating chronic airways disease. There is no evidence on whether or not the ow-dose oxygen deivered by home oxygen concentrators is safe in patients with heart faiure. There is no evidence about the effects of ow-dose oxygen on haemodynamics in patients with severe heart faiure. The equivoca findings are perhaps not surprising. Oxygen is perhaps ikey to be hepfu ony to peope who are hypoxic (i.e. have ow eves of arteria oxygen). Where it has been measured, oxygen has been found to be norma, or even high, during exercise in patients with CHF. 25 When patients with CHF are found to be hypoxic, there is usuay an aternative expanation, such as coincident ung disease or congenita heart disease. 26 Thus, athough patients with CHF may resembe patients with chronic airways disease cinicay, they are much ess ikey to be hypoxic, and so might be thought, a priori, to be ess ikey to gain benefit from oxygen treatment. Seep apnoea One compicating issue in patients with heart faiure is the possibe contribution of seep-disordered breathing (SDB). Depending on the popuation studied, approximatey one-third of patients with heart faiure have SDB. 27 SDB happens when breathing stops during seep. There are two kinds of SDB: obstructive and centra seep apnoea. In obstructive seep apnoea, there is upper airways obstruction from soft tissues; respiratory efforts continue but there is no movement of air into the ungs. The patient usuay wakes and breathing restarts. In centra seep apnoea, the centra drive to breathe stops, usuay in a cycica manner aternating with periods of hyperventiation. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 3

30 INTRODUCTION Patients with SDB, of either kind, become hypoxic during periods of apnoea. It thus might be that patients with heart faiure might benefit from oxygen therapy even in the absence of daytime hypoxia, because it may reieve hypoxia at night. Oxygen therapy might be hepfu for periodic respiration. 28 There are some studies in a sma number of patients investigating the effects of short-term overnight oxygen therapy which suggest that there may be some beneficia effect. For exampe, Toyama et a. 29 studied 20 patients and found improvements in exercise capacity and cardiac function in the 10 patients randomised to overnight oxygen. Other sma studies have found simiar beneficia effects, but there is no systematic review avaiabe. In the argest avaiabe study, Sasayama et a. 30 reported on 52 patients with CHF and a positive overnight seep test randomised to receive NOT or conventiona therapy for 1 year. The group with NOT had better seep patterns and a sight improvement in eft ventricuar ejection fraction (LVEF) and daytime activity eve but no reduction in cardiac events. Home oxygen therapy for breathessness There is surprisingy itte evidence that oxygen is effective in treating breathessness per se. A arge observationa study of patients with a variety of causes of breathessness suggested that oxygen therapy was of no benefit, 31 and apart from the review and a meta-anaysis in peope with COPD discussed in Pumonary disease and oxygen, 15 other systematic reviews identified no evidence that suppementa oxygen heped in the reief of breathessness in patients with heart faiure or patients with breathessness from other causes in the absence of hypoxia A subsequent RCT compared LTOT via concentrator with a sham concentrator for refractory breathessness due to a mixture of aetioogies (mainy COPD or cancer). Athough breathessness improved over 7 days in both groups, neither was superior. The authors suggested that it was simpy airfow over the nasa mucosa, and not oxygen, that might have been the therapeutic intervention. 36 Despite the absence of any evidence in favour of using oxygen, HOT is commony prescribed for severey symptomatic patients with end-stage heart disease: indeed, HOT in the UK is most commony prescribed for conditions other than COPD (Department of Heath, 2011, unpubished data). The Nationa Service Framework for coronary disease specificay recommends considering the potentia benefit from paiative care services and paiation aids (e.g. home oxygen). A Canadian study demonstrated that neary 30% of oxygen prescribing costs were a resut of paiative oxygen prescription (i.e. the patients did not meet the forma criteria for oxygen prescription for respiratory disease). 37 In another survey, 77% paiative physicians gave intractabe dyspnoea as the most common reason for prescription of home oxygen. 38 Fifty per cent of heart faiure patients have breathessness that markedy imits exercise in their ast year of ife, 35 and in an Austraian study between 10% and 20% of patients receiving paiative care were treated with HOT. 39 A sma UK study found that 25% of patients receiving HOT had heart faiure as their underying pathoogy. 40 Finay, it is important to note that data from the Department of Heath suggest that between 24% and 43% of the home oxygen prescribed to approximatey 85,000 patients in Engand is not used or eads to no cinica benefit (Department of Heath, 2011, unpubished data). 4 NIHR Journas Library

31 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 The home oxygen therapy tria rationae Home oxygen therapy is potentiay burdensome for patients and their carers. The concentrator has to be fitted to the home, usuay requiring some driing through was. The device consumes eectricity, athough the costs are met by the NHS. The patient is encumbered to a degree by the oxygen. It is deivered through a nasa cannua, usuay via a ong ength of piping, which imits movement. For some, oxygen cyinders to suppy oxygen when the patient eaves the house are needed. The oxygen can eave the nose and throat feeing very dry. In addition, the stream of oxygen-enriched air is a potentia fire hazard, particuary for patients who continue to smoke. Home oxygen therapy is thus expensive to the heath service and burdensome to the patient, and there is itte evidence of its effectiveness. In the absence of any evidence-based guidance on the use of oxygen therapy for patients with CHF despite its widespread use, there is a need for a tria of HOT to find out if patients gain any benefit from its use. Therefore, the HOT tria was designed to address the question of whether or not there is a roe for HOT in patients with severey symptomatic heart faiure, in terms of its effect on the symptoms for which it is usuay prescribed (e.g. breathessness), QoL and prognosis. We aso panned to conduct a cost-effectiveness anaysis. The study consisted of three parts. The main study was designed to measure the impact of HOT on QoL in severey symptomatic patients. A quaitative substudy was designed to assess the burden on patients and their carers, and an acute oxygen substudy was designed to study whether or not there was any effect on haemodynamics of oxygen given in the same concentration as used by concentrators at home. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 5

32

33 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Chapter 2 Synopsis of tria evoution Tria structure and protoco The tria was designed in response to a ca from the NHS research and deveopment Heath Technoogy Assessment (HTA) programme (HTA number 06/80; see Appendix 1). The grant for the tria was originay awarded in March Stage 1: the North East Oxygen Network tria The HOT tria was originay named the North East Oxygen Network (NEON) tria. The origina conception of the tria was an attempt to provide evidence from a doube-bind tria of the effectiveness of HOT in patients with severey symptomatic CHF. Four centres in the UK were invoved: Hu, Leeds, Darington and Leicester. In the absence of any data from adequatey powered studies, it was impossibe to know what duration of oxygen therapy might be usefu. Whether ong-term therapy, foowing the exampe of successfu treatment for patients with chronic airways disease, or shorter-term overnight oxygen therapy aone was the better treatment was unknown. The study was thus designed in two phases. In the first phase (Figure 1), patients were to be randomised to receive either NOT overnight, assumed to be around 8 hours per night, or LTOT, aiming for at east 15 hours per day. A second randomisation woud aocate patients within the two arms to receive either a true or a sham oxygen concentrator. The design of the second phase depended on the resuts of the first; whichever of NOT or LTOT appeared the better in phase 1 woud then be formay tested in the second phase. Assessed for entry First randomisation NOT LTOT Second randomisation True NOT Sham NOT True LTOT Sham LTOT FIGURE 1 The origina design for the NEON tria. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 7

34 SYNOPSIS OF TRIAL EVOLUTION There was discussion among the research group about the optimum design. Some fet that an open pragmatic tria was best, as this woud estimate treatment effectiveness rather than efficacy, which is an important question for the NHS. On the other hand, a pacebo or sham tria woud estimate the true treatment efficacy of oxygen and ensure binded assessment of outcomes. Indeed, there is some evidence to suggest that the symptom of breathessness is partiay reieved by air bowing over the face. 36,41 Consequenty, the origina tria design was to use sham concentrators for the contro treatment. HOT was to be deivered in the tria using oxygen concentrators rather than oxygen cyinders. A concentrator uses room air and removes a fraction of the nitrogen catayticay, so that the resuting inhaed gas is room air with some nitrogen removed; the concentration of inspired oxygen thus increases from 20.9% to approximatey 28%. The sham machines woud have deivered ony room air. The aim of the piot was to recruit 120 patients, 30 in each of the four potentia arms (see Figure 1). The resuts of the initia piot study were then to inform the second phase study. Whichever of the two oxygen deivery schedues was the more successfu woud be used in a three-way comparison of oxygen therapy, sham therapy and open-abe best medica therapy (BMT). The investigators negotiated with Air Products Heathcare (Air Products and Chemicas, Inc., Aentown, PA, USA), which was abe to manufacture sham machines which deivered room air. At this stage, recruiting sites were thus restricted to those whose contract for home oxygen suppies was with Air Products Heathcare. Pans were deveoped for the two-stage randomisation. The origina grant appication to the HTA programme was with this study design. Probems with the North East Oxygen Network tria Tria management The University of Hu initiay agreed to sponsor the tria; however, after a period of some months it decided not to sponsor the study because of the potentia issue with indemnity which coud arise if a sham oxygen concentrator machine was accidentay given to a patient outside the tria. In addition, a fu-time tria manager had not been incuded in the origina submission to the HTA programme. These probems were rectified by the Hu and East Yorkshire NHS Hospitas Trust agreeing to sponsor the study and through the estabishment of a tria management function. Sham machines A crucia component of the tria, as originay conceived, was the use of sham machines to deiver room air. However, athough it was reativey straightforward for Air Products Heathcare to deiver the sham machines, their use ed to two insuperabe ogistica probems. As the sham machines had to ook as simiar to the rea devices as possibe, it was fet by the company and the sponsor to be a substantia probem that the sham machines might not be kept out of the genera poo of machines intended for use in deivering HOT to non-tria patients. In addition, as the sham machines had to have their aarm functions disabed, there was a risk that they woud not detect eectrica fauts that coud potentiay be a fire hazard. Specific tria insurance for their use was required, but was too costy to aow the tria to proceed. Eventuay, the tria management group accepted that the probems being raised made the tria as originay conceived impossibe to run. Foowing a HTA site monitoring visit by Professor Jenny Hewison and Dr Vaughan Thomas on 24 September 2010, the group redesigned the study and approached the HTA programme to approve an aternative tria design. 8 NIHR Journas Library

35 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Stage 2: the three-arm home oxygen therapy tria The redesigned tria was named the HOT tria. We chose a pragmatic study design avoiding binding and did not incude a sham oxygen arm. In turn, this meant that more oxygen suppy companies coud be used and the number of recruiting centres coud be expanded. An additiona benefit was to avoid the need for specia insurance reated to dummy devices. Furthermore, the answer to a pragmatic tria is arguaby of more reevance to patients and cinicians than the origina study design. The origina research question, however, remained unchanged: to assess the effects of HOT on QoL in patients with CHF (Figure 2). Patients were randomised to receive BMT, BMT pus NOT or BMT pus LTOT for 15 hours per day. The primary end point was QoL at 12 months as measured by the Minnesota Living with Heart Faiure (MLwHF) questionnaire. The tria started recruitment in Apri However, recruitment to the tria was much sower than expected and recruitment targets were not met. There were two major reasons. First, it is a particuar chaenge to recruit patients to studies of paiative care, and the patients being recruited into the HOT study were necessariy very unwe and reaching the end of their ives. Some centres found it difficut to approach such patients and, in some centres, the patients being recruited were predominanty cared for by other groups of heath-care workers in their area, commony in the areas of paiative care, edery care and primary care. Simpy put, frai patients characterised by instabiity needed to be sufficienty fit and stabe to participate in the tria. Assessed for entry Randomisation BMT LTOT NOT n = 150 n = 150 n = 150 FIGURE 2 The origina design for the three-arm HOT tria. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 9

36 SYNOPSIS OF TRIAL EVOLUTION The second, and more intractabe, probem was the time it took to recruit centres to the tria. Mutipe individua site approvas were needed, as for any cinica study, but the pecuiar probem for HOT was the need for approva of HOT. Many sites took the view that this shoud be an NHS excess treatment cost. Hospita trust research and deveopment departments naturay needed approva to prescribe HOT from the reevant primary care providers, eading in many instances to proonged deays. During the tria set-up period of HOT, there were major upheavas in NHS structures. The repacement of primary care trusts with cinica commissioning groups made it extremey chaenging to identify who was responsibe for HOT. In some instances, cinica commissioning groups refused permission for the study to go ahead, and sometimes it proved impossibe to find the responsibe party with whom to negotiate. The median ength of time taken from a centre expressing interest in the study to fina approva was 9 months. Stage 3: the two-arm home oxygen therapy tria The tria management group reaised that, with the difficuties in recruiting both patients and centres, it woud prove impossibe to compete the study within a reasonabe time frame and budget. We again approached the HTA programme to ask permission for a modesty redesigned study. We removed the NOT arm, reducing the study from a three- to a two-armed tria comparing BMT with BMT pus LTOT (15 hours per day). LTOT was chosen over NOT as being the ony form of oxygen therapy shown to improve prognosis, abeit in patients with COPD. The simpification of the study design aowed the sampe size to be greaty reduced (Figure 3). In addition, we brought forward the timing of the primary end point from 12 to 6 months. This had the effect of reducing the time required to foow up patients and increasing the time avaiabe for anaysing the data and writing the report for the HTA. After discussion with the HTA programme, it was agreed that the cost-effectiveness anaysis for the tria was not to be incuded. Assessed for entry Randomisation BMT LTOT FIGURE 3 The fina design for the two-arm HOT tria. n = 111 n = NIHR Journas Library

37 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Chapter 3 The home oxygen therapy tria: methods The HOT tria was a Phase III, prospective, open, pragmatic, muticentred, randomised controed parae-group tria with equa randomisation. The study was designed after extensive consutations with patients in the Department of Academic Cardioogy in Hu. Patients gave very hepfu advice about (1) the design of the study and (2) the format and wording of the patient information eafet. This was particuary usefu, as we were aware of the need to be cear about the requirements of the study, but did not want to cause aarm (particuary about fire risk). The patients aso advised about the number and type of study measures with regard to what was an acceptabe participant burden. We particuary acknowedge the hep and advice we received in designing the study from Mr Patrick Fouk, who generousy agreed to take part in the tria steering committee. Mr Patrick Fouk is an experienced patient representative who has been part of many tria management and tria steering groups. He has been particuary hepfu in ensuring that the tria maintained patient reevance, was not afraid to ask the pertinent, and sometimes difficut, questions during meetings, and gave usefu ongoing advice about recruitment from the patients viewpoint. The aim of the study was to determine whether or not the addition of ong-term HOT (given for at east 15 hours per day) improved the QoL for patients with stabe, severey symptomatic CHF who were aready receiving BMT. Such patients are usuay thought of as being in need of paiative care: their outook is imited and they are severey symptomatic. Any intervention which can reieve symptoms is to be wecomed; conversey, if a treatment is simpy burdensome, its use shoud be abandoned. Participants with NYHA cass III or IV LV systoic dysfunction receiving optima medica therapy were randomised to receive either: BMT pus at east 15 hours LTOT or BMT. The origina tria had a third arm with patients aocated to a NOT-ony group. Twenty-five patients were randomised into the NOT arm before the decision was made to drop the arm in Apri Approvas obtained The Northern and Yorkshire Research Ethics Committee [the Medica Research and Ethics Committee (MREC)] approved the study originay known as the NEON tria on 24 August Further approvas were received on 18 May 2011 and 2 Apri 2013 for changes to the design of the study that were impemented to improve recruitment. The Medicines and Heathcare products Reguatory Agency reviewed the tria protoco and on 28 June 2008 confirmed that oxygen concentrators were medica devices and not medicines. The study, therefore, did not fa under the Cinica Trias Directives and so a cinica tria authorisation was not required. The detais of MREC and research and deveopment department approvas are provided in Appendix 2. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 11

38 THE HOME OXYGEN THERAPY TRIAL: METHODS Patient study group To make the findings of the study as widey appicabe as possibe, the incusion criteria were very broad, with few excusions. Incusion criteria To be incuded in the study, patients had to: be wiing to provide written informed consent and be abe to compete patient assessments be aged 18 years or over have heart faiure from any aetioogy have severe symptoms of heart faiure (NYHA cass III/IV) have LV systoic dysfunction confirmed by echocardiography, with LVEF ess than 40% or graded as at east moderatey impaired on visua inspection if an accurate ejection fraction coud not be cacuated be receiving maximay toerated medica management of their heart faiure as reached target dose of (or be on maximay toerated dose of, or be intoerant of) an inhibitor of the renin angiotensin system shown to improve prognosis reached target dose of (or be on maximay toerated dose of, or be intoerant of) a beta-adrenoceptor antagonist shown to improve prognosis reached target dose of (or be on maximay toerated dose of, or be intoerant of) an adosterone antagonist. Excusion criteria Patients were excuded from the study if they: were unabe to provide written informed consent had had a cardiac resynchronisation therapy device impanted within the previous 3 months had coexisting maignant disease if this woud affect the study in the investigators opinion had interstitia ung disease had COPD ikey to fufi criteria for LTOT; forced expiratory voume in 1 second (FEV 1 )/forced vita capacity (FVC) < 70% and FEV 1 < 40% predicted and hypoxia [partia pressure of arteria oxygen (PaO 2 ) < 7.3kPa or saturations < 90%] were using any device or medication that woud impede their abiity to use LTOT or NOT, such as continuous positive airway pressure were unwiing or unabe to compy with safety reguations regarding oxygen use, particuary smoking were unabe to compete patient-reated information on entry. Randomisation Aocations were centray generated by the York Trias Unit. Patients were initiay randomised into the tria on a 1 : 1 : 1 basis using bock randomisation, with randomy permuted bock sizes of three, six and nine. Subsequenty, after the NOT arm was dropped, 1 : 1 aocation was used, with randomy permuted bock sizes of four, six and eight. Patients were randomised by a member of the research team at the recruiting site using the secure, teephone-based randomisation service at the York Trias Unit. Primary outcome The primary end point was the tota score from the MLwHF questionnaire at 6 months from baseine. As the patient group is highy symptomatic and has a imited prognosis, the 6-month primary outcome is highy cinicay meaningfu. 12 NIHR Journas Library

39 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 The MLwHF questionnaire is a vaidated, disease-specific QoL instrument widey used in heart faiure research to assess both symptom severity and response to treatment. 45,46 It consists of 21 questions focusing on the impact of heart faiure on QoL. Patients are asked to rate the extent to which their heart faiure has prevented them from iving as they wanted during the past month using questions rated on a scae of 0 (no effect) to 5 (very much). The questionnaire is scored by summing the responses to a 21 questions, thus resuting in a score from 0 to 105, with a higher score refecting poorer QoL. The MLwHF questionnaire-vaidated QoL score, version 2, is easy to compete and has been shown to be especiay effective in oder patients with comorbidities. 47 An improvement in the score of 5 is sometimes taken to be a minimum cinicay important difference, 48 but others have suggested that a change of 1 standard error (SE) around the mean score is needed (around 6 or 7, depending on the popuation studied). 49 There are no studies of LTOT to hep guide us. The MIRACLE (Muticenter InSync RAndomized CLinica Evauation) tria of biventricuar pacing was powered for a 13-point improvement in MLwHF questionnaire score and found a score difference of 9 between treatment groups. 50 The CARE-HF (CArdiac REsynchronization-Heart Faiure) study of biventricuar pacing found a score improvement of 10.6 with intervention. 51 In the absence of data on which to base study size, we took a MLwHF questionnaire score of 10 as an arbitrary indicator of the minimum improvement necessary to justify the cost and inconvenience of oxygen therapy for patients. Power cacuation To detect a difference in MLwHF questionnaire score between the two groups of 10 points, with a standard deviation (SD) of 25, at 80% power and 5% significance, we required 100 patients per group. Assuming 10% attrition at 6 months, this equated to 111 per group, a tota sampe size of 222 patients. Secondary outcomes Other indices of QoL, exercise performance and severity of heart faiure were aso measured as part of the study. Measures of quaity of ife Heath status as measured by the European Quaity of Life-5 Dimensions instrument The European Quaity of Life-5 Dimensions (EQ-5D) is a sef-administered, vaidated measure of heath status and consists of a five-question mutiattribute questionnaire and a visua anaogue sef-rating scae. 52,53 Respondents are asked to rate severity of their current probems (eve 1, no probems; eve 2, some/moderate probems; eve 3, severe/extreme probems) for five dimensions of heath: mobiity, sef-care, usua activities, pain/discomfort and anxiety/depression. Patients can be cassified into 243 heath states pus two further additiona states (unconscious and dead). Numerica Rating Scae for breathessness The severity of and distress caused by breathessness was measured by the Numerica Rating Scae (NRS) for breathessness (average and worse over past 24 hours and goba change in breathessness). Patient satisfaction was aso measured by the NRS (in addition to a quaitative substudy to assess patient experience). The NRS measures symptoms and satisfaction on a 10-point scae, anchored at 0 and 10. It is highy correated with visua anaogue scae scores, but is more repeatabe. 31,54 Average and worst breathessness over the past 24 hours were anchored with not breathess and worst breathessness imaginabe. 55 Distress due to breathessness was anchored with no distress at a and the worst imaginabe distress. The minimum cinicay important difference in the scae is 1 point. 56,57 How we a patient had coped with their breathessness over the past 24 hours was anchored on not coping at a and coping very we. Satisfaction with treatment for breathessness was anchored on not at a satisfied and very satisfied. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 13

40 THE HOME OXYGEN THERAPY TRIAL: METHODS Epworth Seepiness Scae score to assess daytime somnoence The Epworth Seepiness Scae (ESS) is a standard scae for screening for, and assessing the severity of, daytime seepiness as part of the SDB syndrome. Patients are asked to rate their chance of dozing in eight different scenarios, such as being a passenger in a car or watching TV. Each is measured on a Likert-type scae from 0, woud never doze, to 3, high chance of dozing, and a tota score is obtained from summing the 8 items out of 24. Mood assessment using the Hospita Anxiety and Depression Scae The Hospita Anxiety and Depression Scae (HADS) is a we-vaidated and easy-to-compete 14-item screening too for depression and anxiety. 58 Each item is measured on a Likert-type scae from 0 to 3, and seven non-overapping items make up each of the two subscaes. For each of the two continuous subscaes (anxiety and depression), patients are categorised as norma (0 7), borderine (8 10) or cinicay anxious/depressed (11 21). Severity of heart faiure The severity of the patient s heart faiure was assessed by measuring LV dysfunction on echocardiography and the N-termina pro-b-type natriuretic hormone (NT-proBNP) eves. Exercise capacity The 6-minute wak test The 6-minute wak test (6MWT) foowed a standardised protoco A fat, obstace-free corridor with chairs paced at either end was used. Patients were instructed to wak as far as possibe, turning 180 every 15 metres in the aotted time of 6 minutes. Patients were abe to rest, if needed, and the time remaining caed every second minute. Patients waked unaccompanied so as not to infuence waking speed. After 6 minutes, patients were instructed to stop and the tota distance covered was measured to the nearest metre. Performance Karnofsky Performance Status scae This vaidated scae incorporates the components of physica activity, work and sef-care of patients. 62,63 Patients are cassified according to their functiona impairment, with the status categories ranging from 0% (dead) to 100% (norma with no compaints and no evidence of disease) in steps of 10%. Athough the Karnofsky Performance Status (KPS) scae is categorica, it is of an ordina nature and was, therefore, treated as continuous data in our anaysis. The scae can be further cassified as died (0); unabe to care for sef; requires equivaent of institutiona or hospita care; disease may be progressing rapidy (10 40); unabe to work; abe to ive at home and care for most persona needs; varying amount of assistance needed (50 70); and abe to carry on norma activity and to work; no specia care needed (80 100). Comorbidity Charson Comorbidity Index This is a vaidated age comorbidity index used to estimate reative risk of death from prognostic cinica covariates, and is usefu in studies with 1 to 2 years foow-up. 64 [See aso cci_js (accessed June 2015).] Prevaence of hypoxia The prevaence of hypoxia was assessed by measuring: resting oxygen saturation oxygen saturation at peak exercise during 6MWT oxygen saturation 5 minutes after the end of the 6MWT nocturna oxygen saturation and the presence of SDB during an overnight seep test using an Embetta (Natus Medica Inc., San Caros, CA, USA). 14 NIHR Journas Library

41 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Safety and adherence We recorded the patients own report of their adherence with the oxygen concentrator, and the number of hours of oxygen used measured by the concentrator meter. Participant death was recorded as a serious adverse event (SAE), with the date of death where possibe. The number of days the patient was aive and out of hospita was cacuated. Other measurements Bood anaysis Standard biochemistry and fu bood count were undertaken. Resuts from these tests undertaken within a month of baseine assessment coud be used. Eectrocardiogram A 12-ead eectrocardiogram was undertaken to determine cardiac rhythm and eectrocardiography intervas. Echocardiogram Routine echocardiographic assessment was performed incuding M-mode, 2-dimensiona images and coour fow Dopper recordings by trained operators. Measurements were taken in accordance with American Echocardiography Society or European Association of Echocardiography guideines. LV systoic function was assessed by attempted measurement of LVEF using Simpson s bipane method in a subjects, and by estimation on a scae of LV systoic impairment as foows: norma, mid, mid to moderate, moderate, moderate to severe or severe systoic impairment. Resuts from echo assessments within 3 months of the baseine assessment were used. New York Heart Association cass Assessment of NYHA grade was made using the foowing cassification: cass I breathess on severe exertion (norma) cass II breathess and/or fatigue on moderate exertion cass III breathess and/or fatigue on mid exertion cass IV breathess and/or fatigue at rest. Spirometry Spirometry was undertaken to determine the FVC and forced expiratory voume in the first 3 seconds. Resuts from spirometry tests undertaken within 3 months of the baseine assessment were used. Other assessments Resting puse rate, bood pressure and respiratory rate, periphera oedema and ung crackes, if any, were measured and recorded. Other data Data recorded incuded height, weight and date of birth, which aowed age at recruitment to be cacuated. Sex, aetioogy of heart faiure, current medication and adverse events were aso recorded. Cost-effectiveness In addition to the EQ-5D instrument, the Heath Service Use Questionnaire was used to measure the eve of heath resource use. Respondents were asked to reca the amount of use they had made of the specified services over the previous 6 months. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 15

42 THE HOME OXYGEN THERAPY TRIAL: METHODS Adverse events In this study an adverse event was defined as any untoward medica occurrence in a patient which did not necessariy have a causa reationship with the study treatments or procedures. Heath-care professionas were asked to report any adverse event occurring in participants in both groups using either the SAE form or the adverse event form within 24 hours of becoming aware of the event. A SAE was defined as an event that resuted in death, was ife-threatening, required hospitaisation or proongation of existing hospitaisation, resuted in persistent or significant disabiity or incapacity, consisted of a congenita anomay or birth defect or was otherwise considered to be medicay significant by the investigator. The reporting heath-care professiona was asked to indicate whether or not, in his or her opinion, the event was reated to the treatment, to indicate if it was expected and to grade the intensity of the event. Further foow-up reports were competed if necessary or unti the oca principa investigator considered the event to be resoved or to have become a chronic ongoing condition. 16 NIHR Journas Library

43 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Chapter 4 Statistica anaysis Anayses mainy compared the LTOT intervention group with the contro (BMT) group on an intention-to-treat basis, incuding a randomised patients in the groups to which they were originay aocated. We performed exporatory anayses incuding contemporaneousy recruited patients in the NOT arm for the primary outcome and surviva anaysis. Anayses were conducted in Stata version 13 (StataCorp LP, Coege Station, TX, USA), using two-sided significance tests at the 5% significance eve. Tria competion The fow of participants through the tria is presented in a CONSORT (Consoidated Standards of Reporting Trias) diagram. The numbers of participants withdrawing from treatment and/or the tria are summarised with the reasons where appicabe (see Figure 6). Baseine data A participant baseine data are summarised descriptivey by treatment group. Continuous measures are reported using summary statistics (mean, SD, median, interquartie range, minimum, maximum) and categorica data are reported as counts and percentages. Comparisons were made between tria groups as randomised and as anaysed in the primary anaysis. No forma statistica comparisons were undertaken. Primary anaysis The primary outcome was heath-reated quaity of ife (HRQoL) as measured by the MLwHF questionnaire scores at 6 months. A ower MLwHF questionnaire score indicates a better QoL. The MLwHF questionnaire consists of 21 items and a tota score is obtained by summing the item scores, where a 21 items have a response. The number of missing MLwHF questionnaire responses was examined at each time point. Mutipe imputation using chained equations was used to fi in missing questionnaire items, where there were fewer than four missing item responses, using other items in the questionnaire ( Linear regression was used to perform the imputation and five imputed data sets were created. The mean of the imputed vaues for each patient was used to repace the missing item. Where the mean was ess than 0 or greater than 5 (the range of permitted scores for the MLwHF questionnaire) it was repaced with the nearest permitted vaue. Our primary anaysis compared MLwHF questionnaire scores between the LTOT and BMT treatment groups using a covariance pattern mixed mode, where effects of interest and baseine covariates are specified as fixed effects, and the correation of observations within patients is modeed by a covariance structure. The outcome modeed was tota MLwHF questionnaire scores at 3, 6 and 12 months. The mode incuded as fixed effects baseine MLwHF questionnaire score, age, og-nt-probnp eve, creatinine eve, treatment group, time and a treatment group time interaction term. Age, NT-proBNP eves and creatinine eves were a continuous variabes as assessed at baseine. NT-proBNP data were found to be significanty positivey skewed and so were og transformed. Different covariance structures for the repeated measurements, which are avaiabe as part of Stata version 13, were expored and the most appropriate pattern used for the fina mode. Diagnostics incuding Akaike s information criterion 65 were compared for each mode (smaer vaues are preferred). Participants were naturay incuded in the mode ony if they had fu data for the baseine covariates and outcome data for at east one post-randomisation time point (3, 6 or 12 months). Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 17

44 STATISTICAL ANALYSIS Estimates of the adjusted mean difference (AMD) between treatment groups in MLwHF questionnaire scores were extracted from the mode for a time points with 95% confidence intervas (CIs) and p-vaues. The primary end point is the treatment effect estimate at 6 months. Estimates for the other time points serve as secondary outcomes. An overa effect of the intervention across a incuded time points was not extracted. Sensitivity anayses Patients were recruited from mutipe centres. To investigate whether or not centre affected the outcome, centre was incuded as a random effect in the primary anaysis mode. 66 Secondary anaysis Comparisons with the nocturna oxygen therapy subgroup Initiay, the HOT tria recruited patients to three tria arms: BMT, LTOT and NOT. In Apri 2013, the decision was made to stop recruitment to the NOT arm, athough patients in this group continued to be foowed up. We conducted an exporatory anaysis on the primary outcome incuding a three treatment arms, which incuded ony contemporaneousy recruited patients, that is ony patients in any arm randomised up to the date at which the randomisation to the NOT arm was cosed. This was to ensure comparabiity of the treatment groups. A covariance pattern mixed mode was used simiar to that described for the primary anaysis incuding a variabe for treatment with the three eves. A non-significant difference was observed between the NOT and LTOT arms in a pairwise comparison from this mode; therefore, these arms were combined and compared against the BMT arm on the primary outcome using a simiar anaysis, and incuding ony patients randomised up to the time that the NOT arm was dropped from the study. Secondary outcomes The foowing outcomes were anaysed using the same method as the primary outcome but just adjusting for the baseine vaue of the dependent variabe: the six questions (separatey) of the NRS for breathessness; ESS; HADS scores for anxiety and depression; KPS of physica activity; metres waked as part of the 6MWT; the Charson Comorbidity Index (CCI); and NT-proBNP eves. Scoring of the secondary outcomes A tota score for the ESS was cacuated ony when a items had a vaid response, in accordance with the scoring instructions detaied at For the HADS, as is recommended, the score for a singe missing item from a subscae was inferred using the mean of the remaining six items. If more than one item was missing, then the subscae was judged as invaid ( A tota score for the CCI was computed by appying a certain number of points for each comorbidity present and adding a score for age ( 40 years, 0 points; years, 1 point; years, 2 points; years, 3 points; and years, 4 points) as detaied at (accessed June 2015). Mortaity Mortaity was anaysed as a time-to-event outcome. For each group, the distribution of time from randomisation to death was described using Kapan Meier surviva estimates. Kapan Meier surviva curves are presented for the two groups. The statistica equivaence of the two curves was tested using the og-rank test. Right censoring occurred if the patient was sti aive at the end of foow-up or if they withdrew or were ost to foow-up. 18 NIHR Journas Library

45 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 We compared the surviva of the LTOT and the BMT groups using a Cox proportiona hazards regression mode adjusting for baseine CCI score. Hazard ratios (HRs) are presented with p-vaues and 95% CIs. Simiar surviva anayses were aso conducted by combining the LTOT and NOT arms, and comparing them with the BMT arm, incuding ony those patients contemporaneousy recruited. Number of days aive and out of hospita The number of days aive and out of hospita (DAOH) was cacuated for each patient as foows: 67 the tota potentia foow-up time was determined as the number of days from randomisation unti the date of the fina foow-up time point (if aive) or the end of study date, 30 May 2014, if the patient had died. Patients who were ost to foow-up had a censoring date of 30 May 2014 to determine potentia foow-up time. The number of nights spent in hospita over the previous 6 months was captured on the patient questionnaire at 6, 12, 18 and 24 months. The tota time spent in hospita was computed by adding the durations of each individua hospita stay. If a patient died, the number of days from their death to the end of the study was assigned as days dead. Days in hospita and days dead were then subtracted from the tota potentia foow-up time to arrive at DAOH for each patient. Summaries of the number and percentage of days the patients in the two groups are aive and out of hospita are presented. Heath-reated quaity of ife as measured by the European Quaity of Life-5 Dimensions Data are summarised for the two treatment groups and a simpe anaysis of variance was used to compare each treatment at each time point. In addition, an unadjusted t-test compared baseine scores at each time point for each treatment group. Prevaence of hypoxia The prevaence of saturation at the threshods of 90% and 95% are summarised for O 2 saturation at rest, at peak and at 5 minutes after the 6MWT. Patient-reported adherence Patient-reported adherence to the oxygen machine is summarised at 3, 6, 12, 18 and 24 months in the LTOT arm and NOT subgroup. Number of hours of oxygen used measured by concentrator meter Summary statistics for the mean number of hours that the machine was used per day by patients in the LTOT arm and NOT subgroup are reported. Adverse events The number of adverse events experienced by each participant and tota number of events overa are summarised for each treatment group. The severity of the event and whether or not it was considered reated to treatment is summarised. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 19

46

47 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Chapter 5 Study procedures Members of the research team participating in the study received good cinica practice training as we as training in a aspects of the tria. Training incuded participant recruitment, eigibiity criteria, tria protoco, adverse event reporting procedures and tria documentation. In order to standardise the study prior to commencement, each study site aso received a tria handbook. Potentia participants for the tria were identified from NHS heart faiure, cardioogy or genera medica cinics. Existing ists of ikey eigibe patients hed within the NHS hospitas were aso reviewed. The study was introduced to patients by the cinician responsibe for their treatment (usuay a consutant cardioogist). Patients then had the opportunity to discuss study participation with the research nurse. In order to aid recruitment some sites used patient identification centres. Potentia participants were sent an introduction etter with an invitation to contact the study team if they were interested in taking part in the study. Aternativey, the research nurse coud contact the patient directy by teephone. Each potentia participant was informed of the aims, methods, expected benefits, potentia hazards and discomforts of the study verbay and through a patient information sheet (see Appendix 3). Participants were given at east 24 hours to consider participation in the study. Participants who wished to take part in the study provided written informed consent (see Appendix 4). Baseine data were then coected. Each participant s genera practitioner was notified of the patient s invovement in the HOT tria and their group aocation after recruitment. The fow of participants through the tria is presented in a CONSORT 68 diagram (see Figure 6). Baseine and foow-up assessments After written informed consent had been obtained, baseine data were coected using the nurse and patient baseine questionnaires. After giving consent and competing baseine assessments, patients were randomised by a member of the research team at the recruiting site using the secure, teephone-based randomisation service at the York Trias Unit. In the event that the oca cinica team fet that a patient needed to be assessed for obstructive seep apnoea, the patient was not randomised unti a cinica decision was made. Home oxygen therapy If the patient was randomised to receive HOT, a cinica prescription was competed and sent to the oca oxygen suppy company hoding the standard NHS contract within that particuar region. These were Air Liquide UK (Birmingham, UK), Doby Viviso (Stiring, UK), BOC Heathcare (Manchester, UK) and Air Products (Crewe, UK). Concentrators were deivered to the patient s home by the recruiting hospita s usua oxygen suppier, in accordance with existing NHS agreements. The oxygen suppy company typicay instaed the equipment within 3 days of the prescription being issued. The instaing engineer instructed the participant about the safety requirements of using the machines and gave detais of how to caim for the eectricity costs of running the machine. In accordance with NHS agreements, the concentrators shoud have been serviced approximatey every 6 months. At the end of an individua patient s tria participation, oxygen concentrators were removed from the patients homes uness they wished to continue treatment. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 21

48 STUDY PROCEDURES Best medica therapy Patients aocated to this arm received the BMT currenty avaiabe. Patients received the maximay toerated medica management for their heart faiure and reached their target dose of inhibitors of the renin angiotensin system, a beta-adrenoceptor antagonist and an adosterone antagonist. Foow-up Patients with CHF usuay attend cinic every 6 months and foow-up in the study was arranged around standard attendances to prevent patients being unduy burdened with additiona hospita visits. At the concusion of the tria, fina cinica data were coected using existing hospita records, incuding admission and mortaity. The data coected are summarised in Tabe 2, together with the schedue for data coection. TABLE 2 Schedue of patient assessments: study structure Months after recruitment Assessment Baseine: cinic 3: home or cinic 6: cinic 12: cinic 18: cinic 24: cinic Cinica Age, sex, aetioogy, height Weight Current medication Resting puse rate and bood pressure, respiratory rate Assessment of periphera oedema and ung crackes ECG Bood test BCP, FBC (standard biochemistry and haematoogy) Bood test NT-proBNP Spirometry Echocardiogram 6MWT and pre/post O 2 saturation Overnight seep test (if ocay accessibe) CCI KPS score QoL MLwHF questionnaire NRS breathessness HADS ESS Heath economics EQ-5D Heath Service Use Questionnaire (not a questions) BCP, biochemica profie; ECG, eectrocardiography; FBC, fu bood count. 22 NIHR Journas Library

49 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Tria competion Participants were deemed to have exited the tria when: they had been in the tria for 24 months or 6 months if there was insufficient time to foow the patient further they wished to withdraw from the tria their treating physician or medica researcher withdrew them from the tria they were ost to foow-up they died. As we as withdrawing fuy from the tria, participants had the option of: withdrawing from receiving oxygen (if that had been their aocation) withdrawing from the coection of data via patient questionnaires withdrawing from the coection of data via nurse questionnaires. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 23

50

51 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Chapter 6 Resuts Tria recruitment The HOT study was stopped eary by the funders because of poor patient adherence to the oxygen prescription. Recruitment began in Apri 2012 and stopped in February Randomisation to the NOT arm was stopped in Apri In tota, 139 patients were randomised, 57 to each of the LTOT and BMT arms and 25 to the NOT arm. The overa rate of recruitment is shown in Figure 4. The origina sampe size for the three-arm tria was 450 patients. The aim was to recruit these patients in 12 months; however, recruitment was sower than expected and, by the end of Apri 2013, 74 patients had been recruited into the tria (25 to each of the LTOT and NOT arms and 24 to the BMT arm). The decision was made to drop the NOT arm and continue the tria with two arms with a revised sampe size of 222 patients. The recruitment period was extended to August When the tria was cosed at the end of February 2014, a tota of 139 participants had been randomised. Over the course of the tria, a tota of 15 participating sites joined, a in the UK (Tabe 3). Recruitment was staggered, with sites joining over the course of the tria. At east one tria participant was recruited in 13 out of the 15 sites (Figure 5 and Tabe 3). The median number of participants recruited per site was 4 (range 1 76). Over haf of the participants (n = 76) were recruited from the Hu site, where the chief investigator was based. Two sites did not recruit any patients. Significant attempts were made to recruit a patient in Aneurin Bevan University Heath Board but no eigibe, consenting patients were identified; East Cheshire NHS Trust was granted research and deveopment approva ony shorty before recruitment ceased and, therefore, did not have time to recruit a patient. Figure 6 shows the CONSORT fow diagram of participants through the tria. 500 Number of participants NOT arm dropped Target (three arm) Target (two arm) Actua (three arm) Actua (two arm) 0 Apri 2012 June 2012 August 2012 October 2012 December 2012 February 2013 Apri 2013 June 2013 August 2013 October 2013 December 2013 February 2014 FIGURE 4 The overa rate of recruitment in the HOT tria. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 25

52 RESULTS TABLE 3 Number of participants randomised by group and site, n (%) Site Principa investigator Started recruiting LTOT (n = 57) NOT (n = 25) BMT (n = 57) Tota (n = 139) Hu Professor Andrew Cark Apri (45.6) 18 (72.0) 32 (56.1) 76 (54.7) Chesterfied Dr Justin Cooke October (12.3) 0 (0.0) 5 (8.8) 12 (8.6) Odham Dr Joanta Soboewska January (10.5) 3 (12.0) 3 (5.3) 12 (8.6) Darington Professor Jerry Murphy September (7.0) 0 (0.0) 4 (7.0) 8 (5.8) Dundee Dr Mies Witham November (10.5) 1 (4.0) 0 (0.0) 7 (5.0) Leicester Professor Iain Squire November (3.5) 0 (0.0) 5 (8.8) 7 (5.0) Barnet Dr Ameet Bakhai October (1.7) 2 (8.0) 1 (1.8) 4 (2.9) Durham Dr Mohamed E-Harari January (0.0) 1 (4.0) 3 (5.3) 4 (2.9) Bradford Dr Pau Smith January (3.5) 0 (0.0) 1 (1.8) 3 (2.2) Eaing Dr Stuart Rosen May (0.0) 0 (0.0) 2 (3.5) 2 (1.4) Sunderand Dr John Baxter June (1.8) 0 (0.0) 1 (1.7) 2 (1.4) Pinderfieds Dr Pau Brooksby Juy (1.8) 0 (0.0) 0 (0.0) 1 (0.7) Pymouth a Dr Andrew Stone June (1.8) 0 (0.0) 0 (0.0) 1 (0.7) a Pymouth started recruitment after the NOT arm was dropped and so coud not randomise a patient to the NOT arm Number of participants Hu Odham Chesterfied Darington Leicester Dundee Site Durham Barnet Bradford Sunderand Eaing Pymouth Pinderfieds FIGURE 5 Participant recruitment by site (two sites did not recruit any patients). 26 NIHR Journas Library

53 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Randomised (n = 139) LTOT (n = 57) BMT (n = 57) NOT (n = 25) Nurse questionnaires Month 3, n = 54 Month 6, n = 45 Month 12, n = 22 Patient questionnaires Month 3, n = 53 Month 6, n = 45 Month 12, n = 22 Lost to foow-up up to 12 months Died, n = 6 Withdrew from tria, n = 2 Primary outcome anaysis a (n = 51) Nurse questionnaires Month 3, n = 54 Month 6, n = 45 Month 12, n = 18 Patient questionnaires Lost to foow-up up to 12 months Primary outcome anaysis a (n = 51) Nurse questionnaires Month 3, n = 20 Month 6, n = 17 Month 12, n = 17 Month 18, n = 9 Month 24, n = 2 Month 3, n = 53 Month 6, n = 43 Month 12, n = 18 Month 3, n = 20 Died, n = 11 Withdrew from tria, n = 1 Patient questionnaires Month 6, n = 17 Month 12, n = 17 Month 18, n = 9 Month 24, n = 2 Lost to foow-up Died, n = 7 Withdrew from tria, n = 2 Nurse questionnaires Month 18, n = 9 Month 24, n = 2 Patient questionnaires Month 18, n = 8 Month 24, n = 2 Lost to foow-up up to 24 months Died, n = 0 Withdrew from tria, n = 1 Nurse questionnaires Month 18, n = 9 Month 24, n = 1 Patient questionnaires Month 18, n = 8 Month 24, n = 1 Lost to foow-up up to 24 months Died, n = 1 Withdrew from tria, n = 0 Recruitment into the NOT arm ceased after the randomisation of 74 participants (LTOT n = 25; BMT n = 24; NOT n = 25). Participants in the NOT arm continued to be foowed up but were not incuded in the main tria anaysis FIGURE 6 The CONSORT fow diagram. a, Based on patients who provided fu covariate data for the primary anaysis mode and primary outcome data at at east one of the 3-, 6- or 12-month time points. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 27

54 RESULTS Withdrawas A greater proportion of patients in the LTOT arm (n = 8, 14%) than the NOT arm (n = 2, 8%) formay withdrew from their aocated tria treatment. One patient in the NOT arm withdrew from competing the posta patient questionnaires and one patient in each of the NOT and BMT arms withdrew from competing the nurse questionnaires. A tota of four patients (one LTOT, one BMT and two NOT) requested fu tria withdrawa and two further patients (one LTOT, one NOT) were withdrawn by a heath professiona. The reasons given for each of the change of circumstances are shown in Tabe 4. There were 25 recorded deaths during the course of the tria (LTOT, 6; BMT, 12; and NOT, 7). Baseine participant characteristics Competed baseine questionnaires were received from 139 (100%) randomised participants. Participant baseine characteristics are summarised by treatment group (LTOT and BMT arms ony) in Tabes 5 and 6. The majority of patients in the study were mae (n = 80, 70%) and the mean age was 72 years (range years). The most common cause of the participants heart faiure was ischaemic or coronary heart disease (n = 96, 84%), and the vast majority of participants were in NHYA cass III (n = 108, 95%). TABLE 4 Reasons (where given) for participant change of circumstances Reason LTOT, n (%) NOT, n (%) BMT, n (%) Overa, n (%) Withdrew from treatment n = 8 n = 2 n = 10 Did not fee oxygen was heping 3 (37.5) 0 (0) 3 (30) Was not using oxygen 2 (25.0) 1 (50) 3 (30) Probems seeping/at night 0 (0) 1 (50) 1 (10) Worried about tripping over tubing 1 (12.5) 0 (0) 1 (10) Cannua and mask uncomfortabe 1 (12.5) 0 (0) 1 (10) No reason given 1 (12.5) 0 (0) 1 (10) Withdrew from patient questionnaires n = 0 n = 1 n = 0 n = 1 No reason given 0 (0) 1 (100) 0 (0) 1 (100) Withdrew from nurse questionnaires n = 0 n = 1 n = 1 n = 2 Did not wish to attend hospita for visits 0 (0) 1 (100) 1 (100) 2 (100) Withdrew from tria n = 1 n = 2 n = 1 n = 4 Did not fee study was beneficia 1 (100) 0 (0) 0 (0) 1 (25) Did not want assessments/site visits 0 (0) 0 (0) 1 (100) 1 (25) Had not used oxygen at a as was afraid of it, and did not want to be foowed up 0 (0) 1 (50) 0 (0) 1 (25) No reason given 0 (0) 1 (50) 0 (0) 1 (25) Withdrawn by heath-care professiona n = 1 n = 1 n = 0 n = 2 Patient in hospice and unwe 0 (0) 1 (100) 0 (0) 1 (50) Patient incapacitated due to stroke 1 (100) 0 (0) 0 (0) 1 (50) 28 NIHR Journas Library

55 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 TABLE 5 Baseine participant characteristics by treatment group Characteristic LTOT (n = 57) BMT (n = 57) Tota (n = 114) Sex, n (%) Mae 40 (70.2) 40 (70.2) 80 (70.2) Femae 17 (29.8) 17 (29.8) 34 (29.8) Age (years), n (%) Mean (SD) 73.1 (10.6) 71.4 (11.9) 72.3 (11.3) Median (min., max.) 74.7 (51.7, 94.0) 74.4 (38.9, 87.4) 74.7 (38.9, 94.0) Height (m) Mean (SD) 1.68 (0.10) 1.68 (0.08) 1.68 (0.09) Median (min., max.) 1.67 (1.43, 1.85) 1.67 (1.48, 1.90) 1.67 (1.43, 1.90) Weight (kg) Mean (SD) 83.0 (20.7) 84.1 (21.8) 83.5 (21.2) Median (min., max.) 83.2 (41.5, 165.1) 82.0 (50.0, 140.0) 82.1 (41.5, 165.1) Aetioogy, n (%) IHD 51 (89.5) 45 (79.0) 96 (84.2) Arrhythmia 15 (26.3) 21 (36.8) 36 (31.6) Diated cardiomyopathy 4 (7.0) 10 (17.5) 14 (12.3) Vavuar heart disease 2 (3.5) 3 (5.3) 5 (4.4) Amyoidosis 0 (0.0) 1 (1.8) 1 (0.9) Hypertension 0 (0.0) 1 (1.8) 1 (0.9) NYHA cass, n (%) III 57 (100.0) 51 (89.5) 108 (94.7) IV 0 (0.0) 6 (10.5) 6 (5.3) NT-proBNP (ng/) Mean (SD) (8402.0) (4026.7) (6627.3) Median (min., max.) 2243 (118, 35,000) 1931 (82, 15,594) (82, 35,000) Creatinine (µmo/) Mean (SD) (44.1) (50.2) (47.1) Median (min., max.) 113 (66, 252) (63, 235) 114 (63, 252) LV ejection fraction (%) Mean (SD) 28.0 (7.7) 28.2 (8.1) 28.1 (7.9) Median (min., max.) 29 (7, 39) 28 (11, 50) 28 (7, 50) LV impairment, n (%) Mid to moderate 1 (1.8) 0 (0.0) 1 (0.9) Moderate 11 (19.3) 9 (15.8) 20 (17.5) Moderate to severe 10 (17.6) 12 (21.1) 22 (19.3) Severe 35 (61.4) 36 (63.2) 71 (62.3) ACE inhibitors/arbs, n (%) 50 (87.7) 48 (84.2) 98 (86.0) Beta-bockers, n (%) 53 (93.0) 49 (86.0) 102 (89.5) Adosterone antagonists, n (%) 47 (82.5) 38 (66.7) 85 (74.6) ACE, angiotensin-converting enzyme; ARB, angiotensin receptor bocker; IHD, ischaemic heart disease; max., maximum; min., minimum. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 29

56 RESULTS TABLE 6 Baseine characteristics by treatment group Characteristic LTOT (n = 57) BMT (n = 57) Tota (n = 114) Puse rate (beats per minute) Mean (SD) 68.7 (13.0) 70.5 (12.8) 69.6 (12.9) Median (min., max.) 69 (18, 99) 72 (15, 97) 70 (15, 99) Systoic bood pressure (mmhg) Mean (SD) (20.8) (20.8) (20.8) Median (min., max.) 117 (78, 180) 114 (65, 194) 115 (65, 194) Diastoic bood pressure (mmhg) Mean (SD) 65.1 (9.8) 65.1 (13.3) 65.1 (11.6) Median (min., max.) 63 (46, 100) 63 (25, 111) 63 (25, 111) Respiratory rate (breaths per minute) Mean (SD) 16.6 (4.7) 15.6 (3.4) 16.1 (4.1) Median (min., max.) 16 (10, 36) 16 (9, 24) 16 (9, 36) FVC () Mean (SD) 2.4 (0.8) 2.6 (0.9) 2.5 (0.9) Median (min., max.) 2.34 (0.89, 4.42) 2.48 (0.73, 5.28) 2.37 (0.73, 5.28) FEV 1 () Mean (SD) 1.7 (0.8) 1.8 (0.7) 1.8 (0.7) Median (min., max.) 1.65 (0.55, 3.49) 1.78 (0.45, 3.80) 1.74 (0.45, 3.80) Rhythm, n (%) Sinus rhythm 29 (50.9) 25 (43.9) 54 (47.4) Atria fibriation 14 (24.6) 16 (28.1) 30 (26.3) Atria futter 0 (0.0) 0 (0.0) 0 (0.0) Biventricuar pacing synchronous 12 (21.1) 12 (21.1) 24 (21.1) Biventricuar pacing asynchronous 0 (0.0) 3 (5.3) 3 (2.6) RV pacing synchronous 1 (1.8) 0 (0.0) 1 (0.9) RV pacing asynchronous 1 (1.8) 0 (0.0) 1 (0.9) Other 7 (12.3) 3 (5.3) 10 (8.8) Haemogobin (g/) Mean (SD) (16.7) (19.6) (18.2) Median (min., max.) 128 (96, 170) 127 (92, 179) (92, 179) max., maximum; min., minimum; RV, right ventricuar. 30 NIHR Journas Library

57 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 To be eigibe for the tria, participants had to have a LVEF < 40% or be graded as at east moderatey impaired on visua inspection. LVEF was not recorded in eight participants in the LTOT or BMT arm, and a of these participants had either moderate or severe LV impairment. One participant with a LVEF > 40% was randomised as he or she was visuay assessed as having severe LV impairment. NT-proBNP eve was not an entry criterion to the study, but the very high eves suggest that patients with severe heart faiure were recruited. In genera, the two treatment groups were comparabe across baseine participant characteristics; however, there was a sight imbaance in NT-proBNP eve and proportion of patients taking adosterone antagonists. The mean NT-proBNP eve was higher in the LTOT arm than in the BMT arm, and the proportion of patients taking an adosterone antagonist was greater in the LTOT arm. NT-proBNP eve was pre-specified as a covariate in the primary anaysis and so this imbaance was controed for. Primary outcome The primary outcome was MLwHF questionnaire score at 6 months post randomisation. At baseine, a response to one item was missing in two patients and responses to two items were missing in two patients. At 3 months, among those for whom the MLwHF questionnaire was returned, there were no missing data. At 6 months, seven patients had a missing response to one item. At 12 months, one questionnaire was returned not competed, with a note to say that the patient was too unwe to compete the questionnaire; otherwise there were no missing data. At 12, 18 and 24 months, where patients returned a questionnaire, there were compete item data for the MLwHF questionnaire. Imputation of missing items was, therefore, necessary ony on baseine and 6-month data. Summaries of the MLwHF questionnaire score (post imputation) by treatment group at each time point are presented in Tabe 7. A covariance pattern mode was used to compare MLwHF questionnaire score between the LTOT and BMT arms, adjusting for baseine MLwHF questionnaire score, age, og-nt-probnp eve, creatinine eve, treatment group, time and a treatment group time interaction. The mode incuded a patients who provided fu data for each of the incuded covariates, and MLwHF questionnaire outcome data at one or more post-baseine time points up to 12 months, and so was based on 102 out of 114 patients (89.5%): 51 (89.5%) in each group. Baseine characteristics of participants as incuded in primary anaysis mode are compared between the treatment arms in Tabe 8. It does not appear that the oss of the 12 patients for whom covariate or outcome data were missing significanty impacted on the baance between the treatment arms achieved at randomisation. The assumptions of the inear mode were checked visuay. The normaity of the standardised residuas was assessed via a histogram and Q Q pot (see Appendix 5), and the homoscedasticity of the errors was checked by potting the residuas against the fitted vaues. These pots gave no reason to be concerned about the vaidity of the assumptions. In tota, 88 participants provided vaid MLwHF questionnaire data at 6 months (LTOT n = 45; BMT n = 43); however, five of these participants did not provide data for at east one of the incuded baseine covariates (MLwHF questionnaire score, age, NT-proBNP eve, creatinine eve) so they were not incuded in the mode. A further 19 participants (LTOT n = 8; BMT n = 11) were incuded in the mode, as they provided vaid MLwHF questionnaire data at 3 and/or 12 months, resuting in an anaysed sampe of 102 participants. An estimate of the treatment effect at 6 months was extracted from the mode. As we as the 85 participants in the mode who provided primary outcome data at 6 months, participants who provided data at 3 and/or 12 months but not at 6 months are taken into account when the treatment effect at 6 months is estimated owing to the specification of the covariance pattern between the within-patient repeated measures. There was no evidence of a difference in MLwHF questionnaire score between the LTOT and BMT groups at 6 months (AMD 0.10, 95% CI 6.88 to 6.69; p = 0.98) (see Tabe 7); the LTOT group had a sighty ower MLwHF questionnaire score at 6 months, but the difference was not statisticay significant. Figure 7 pots the adjusted means obtained from the mode by treatment group over time. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 31

58 RESULTS TABLE 7 Raw and adjusted primary outcome data Unadjusted Adjusted LTOT (n = 57) BMT (n = 57) Tota (n = 114) LTOT BMT Time point n Mean (SD) Median (min., max.) n Mean (SD) Median (min., max.) n Mean (SD) Median (min., max.) Mean (SE) 95% CI Mean (SE) 95% CI Mean difference (95% CI); p-vaue Baseine (18.4) 52 (13, 95) (17.9) 50 (20, 93) (18.1) 52 (13, 95) Month (16.2) 46 (7, 84) (18.2) 52 (2, 99) (19.2) 48 (2, 91) 46.5 (1.8) 42.9 to (1.8) 48.4 to ( 10.5 to 0.4); p = 0.03 Month (18.5) 49 (2, 81) (20.2) 47 (10, 91) (19.2) 48 (2, 91) 49.2 (2.4) 44.5 to (2.5) 44.5 to ( 6.9 to 6.7); p = 0.98 Month (16.0) 46 (20, 79) (18.8) 51 (11, 77) (17.1) 49 (11, 79) 46.5 (2.8) 40.9 to (3.2) 42.9 to ( 11.0 to 5.8); p = 0.54 Month (18.6) 35.5 (25, 72) (20.4) 30.5 (12, 72) (19.2) 33 (12, 72) Month (13.4) 43.5 (34, 53) (10.4) 51 (34, 53) max., maximum; min., minimum. 32 NIHR Journas Library

59 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 TABLE 8 Baseine characteristics of participants as incuded in the primary anaysis by treatment group Characteristic LTOT (n = 51) BMT (n = 51) Tota (n = 102) Sex, n (%) Mae 35 (68.6) 36 (70.6) 71 (69.6) Femae 16 (31.4) 15 (29.4) 31 (30.4) Age (years) Mean (SD) 72.0 (9.8) 71.5 (11.9) 71.8 (10.9) Median (min., max.) 74.1 (51.7, 91.8) 74.9 (38.9, 86.4) 74.5 (38.9, 91.8) Height (m) Mean (SD) 1.68 (0.10) 1.68 (0.08) 1.68 (0.09) Median (min., max.) 1.68 (1.43, 1.85) 1.67 (1.48, 1.90) 1.67 (1.43, 1.90) Weight (kg) Mean (SD) 83.1 (20.9) 83.8 (21.9) 83.5 (21.3) Median (min., max.) 84.2 (41.5, 165.1) 82.0 (50.0, 140.0) 82.4 (41.5, 165.1) Aetioogy, n (%) IHD 47 (92.2) 39 (76.5) 86 (84.3) Arrhythmia 14 (27.5) 17 (33.3) 31 (30.4) Diated cardiomyopathy 4 (7.8) 10 (19.6) 14 (13.7) Vavuar heart disease 1 (2.0) 3 (5.9) 4 (3.9) Amyoidosis 0 (0.0) 1 (2.0) 1 (1.0) Hypertension 0 (0.0) 1 (2.0) 1 (1.0) NYHA cass, n (%) III 51 (100.0) 45 (88.2) 96 (94.1) IV 0 (0.0) 6 (11.8) 6 (5.9) NT-proBNP (ng/) Mean (SD) (7288.2) (4081.2) (5898.9) Median (min., max.) 2198 (118, 35,000) 1900 (82, 15,594) (82, 35,000) Creatinine (µmo/) Mean (SD) (46.1) (51.6) (48.9) Median (min., max.) 112 (66, 252) 127 (63, 235) (63, 252) LV ejection fraction (%) Mean (SD) 28.5 (7.8) 28.7 (8.3) 28.6 (8.0) Median (min., max.) 30 (7, 39) 29 (11, 50) 29.5 (7, 50) LV impairment, n (%) Mid to moderate 1 (2.0) 0 (0.0) 1 (1.0) Moderate 10 (19.6) 9 (17.7) 19 (18.6) Moderate to severe 10 (19.6) 11 (21.6) 21 (20.6) Severe 30 (58.8) 31 (60.8) 61 (59.8) ACE inhibitors/arbs, n (%) 44 (86.3) 43 (84.3) 87 (85.3) Beta-bockers, n (%) 47 (92.2) 43 (84.3) 90 (88.2) Adosterone antagonists, n (%) 41 (80.4) 35 (68.6) 76 (74.5) ACE, angiotensin-converting enzyme; ARB, angiotensin receptor bocker; IHD, ischaemic heart disease; max., maximum; min., minimum. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 33

60 RESULTS Mean MLwHF score by group BMT LTOT 5 0 Baseine Month 3 Month 6 Month 12 Time of foow-up FIGURE 7 Adjusted means for MLwHF questionnaire score by treatment group over time from the primary anaysis mode. Predictors of Minnesota Living with Heart Faiure questionnaire score at 6 months Tabe 9 presents the coefficients from the primary anaysis mode. The baseine MLwHF questionnaire score, creatinine eve and og-nt-probnp eve were significant predictors of the outcome MLwHF questionnaire score. TABLE 9 Coefficients from the primary anaysis mode Variabe Coefficient SE 95% CI p-vaue Baseine MLwHF questionnaire score to 0.73 < Age to Creatinine to Log-NT-proBNP to Aocation LTOT to Time point 6 months to months to Aocation (time point) LTOT (6 months) to LTOT (12 months) to Constant to NIHR Journas Library

61 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Minnesota Living with Heart Faiure questionnaire score at 3 and 12 months From the primary anaysis mode, the mean difference between groups for the outcome at 3 and 12 months was extracted. At 3 months, there was evidence to suggest that LTOT patients had a ower MLwHF questionnaire score (AMD 5.47, 95% CI to 0.41; p = 0.03); however, this difference did not persist to 6 or 12 months (12-month AMD 2.64, 95% CI to 5.75; p = 0.54) (see Tabe 3). Minnesota Living with Heart Faiure questionnaire score at 6 months adjusting for centre Anaysis was undertaken to adjust for possibe correation between MLwHF questionnaire score within a centre, by incuding centre as a random effect in the primary anaysis mode. No evidence of a difference between the LTOT and BMT arms was found at 6 months (AMD 0.21, 95% CI 6.98 to 6.57; p = 0.95) (see Tabe 7). Note that an unstructured covariance pattern was used for a anayses of the MLwHF questionnaire scores. The nocturna oxygen therapy subgroup Patients were recruited to the NOT arm of the tria unti 30 Apri The baseine participant characteristics of a patients randomised up to this date are shown in Tabe 10 by treatment group. We conducted an exporatory anaysis on the primary outcome incuding a three treatment arms, which incuded ony contemporaneousy recruited patients. That is, patients aocated to the NOT arm were compared against other patients randomised up to 30 Apri This was to ensure comparabiity of the treatment groups. A covariance pattern mixed mode (unstructured correation) was used to compare MLwHF questionnaire scores between the three treatment groups. The outcome modeed was tota MLwHF questionnaire scores up to 12 months, and the mode was adjusted for baseine MLwHF questionnaire score, age, creatinine eve, og-nt-probnp eve, treatment group, time and a treatment group time interaction. The overa treatment effect was not found to be significant [χ 2 = 2.94, degrees of freedom (df) = 2; p = 0.23] so no pairwise comparisons were empoyed. As there was a non-significant difference between the NOT and LTOT arms (contrast 5.28, 95% CI to 3.60; p = 0.24), these arms were combined and compared against the BMT arm on the primary outcome using an anaysis simiar to that described for the primary anaysis and incuding ony patients randomised up to the time that the NOT arm was dropped from the study. There was no evidence of a difference between the combined HOT groups and the BMT group at 3, 6 or 12 months (Tabe 11). Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 35

62 RESULTS TABLE 10 Baseine participant characteristics of a patients randomised up to 30 Apri 2013 by treatment group Characteristic LTOT (n = 25) NOT (n = 25) BMT (n = 24) Tota (n = 74) Sex, n (%) Mae 17 (68.0) 22 (88.0) 17 (70.8) 56 (75.7) Femae 8 (32.0) 3 (12.0) 7 (29.2) 18 (24.3) Age (years) Mean (SD) 70.7 (10.9) 75.8 (9.1) 71.6 (11.2) 72.7 (10.5) Median (min., max.) 73.7 (51.7, 89.7) 77.4 (60.3, 89.9) 73.9 (38.9, 84.5) 74.5 (38.9, 89.9) Height (m) Mean (SD) 1.69 (0.09) 1.71 (0.08) 1.70 (0.09) 1.70 (0.08) Median (min., max.) 1.65 (1.52, 1.82) 1.71 (1.54, 1.83) 1.69 (1.49, 1.90) 1.70 (1.49, 1.90) Weight (kg) Mean (SD) 88.5 (21.6) 80.4 (14.6) 86.9 (22.6) 85.2 (19.9) Median (min., max.) 85.5 (56.2, 165.1) 79.3 (49.6, 105.6) 85.5 (50.0, 128.0) 84.6 (49.6, 165.1) Aetioogy, n (%) IHD 23 (92.0) 22 (88.0) 17 (70.8) 62 (83.8) Arrhythmia 7 (28.0) 11 (44.0) 9 (37.5) 27 (36.5) Diated cardiomyopathy 2 (8.0) 3 (12.0) 5 (20.8) 10 (13.5) Vavuar heart disease 0 (0.0) 1 (4.0) 3 (12.5) 4 (5.4) NYHA cass, n (%) III 25 (100.0) 17 (68.0) 21 (87.5) 63 (85.1) IV 0 (0.0) 8 (32.0) 3 (12.5) 11 (14.9) NT-proBNP (ng/) Mean (SD) (6738.0) (7223.5) (4817.6) (6342.7) Median (min., max.) 1662 (118, 28,504) 3276 (329, 25,674) (82, 15,594) 2177 (82, 28,504) Creatinine (µmo/) Mean (SD) (46.7) (43.7) (50.4) (46.9) Median (min., max.) 112 (67, 213) 148 (83, 235) (70, 235) 132 (67, 235) LV ejection fraction (%) Mean (SD) 27.7 (6.9) 27.8 (6.4) 29.4 (6.1) 28.3 (6.4) Median (min., max.) 29 (7, 37.5) 29.5 (15, 40) 29.5 (18, 40) 29 (7, 40) LV impairment, n (%) Mid to moderate 1 (4.0) 0 (0.0) 0 (0.0) 1 (1.4) Moderate 5 (20.0) 3 (12.0) 5 (20.8) 13 (17.6) Moderate to severe 5 (20.0) 6 (24.0) 5 (20.8) 16 (21.6) Severe 14 (56.0) 16 (64.0) 14 (58.3) 44 (59.5) ACE inhibitors/arbs, n (%) 22 (88.0) 21 (84.0) 22 (91.7) 65 (87.8) Beta-bockers, n (%) 25 (100.0) 22 (88.0) 22 (91.7) 69 (93.2) Adosterone antagonists, n (%) 18 (72.0) 18 (72.0) 16 (66.7) 52 (70.3) ACE, angiotensin-converting enzyme; ARB, angiotensin receptor bocker; IHD, ischaemic heart disease; max., maximum; min., minimum. 36 NIHR Journas Library

63 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 TABLE 11 Observed and adjusted summaries for the primary outcome measure by treatment group at each time point Unadjusted Adjusted LTOT (n = 25) NOT (n = 25) BMT (n = 24) Tota (n = 74) LTOT + NOT BMT Time point n Mean (SD) Median (min., max.) n Mean (SD) Median (min., max.) n Mean (SD) Median (min., max.) n Mean (SD) Median (min., max.) Mean (SE) 95% CI Mean (SE) 95% CI Mean difference (95% CI); p-vaue Baseine (18.5) 58 (23, 95) (16.5) 57 (35, 90) (18.5) 55.5 (21, 93) (17.7) 57 (21, 95) Month (17.2) 52 (7, 84) (20.8) 56.5 (7, 83) (18.4) 51 (2, 86) (18.6) 52 (2, 86) 9.7 (2.4) 52.7 (3.4) 3.0 ( 11.0 to 5.0); p = 0.46 Month (19.2) 49 (11, 81) (20.9) 37 (0, 80) (23.6) 46 (10, 91) (21.4) 45 (0, 91) 45.8 (3.0) 39.9 to (4.3) 41.6 to ( 14.7 to 6.1); p = 0.42 Month (16.4) 46 (20, 79) (15.3) 37 (12, 66) (19.4) 51 (11, 77) (17.3) 42 (11, 79) 43.6 (2.7) 38.4 to (4.2) 42.6 to ( 16.8 to 2.6); p = 0.15 Month (18.6) 35.5 (25, 72) (20.7) 37 (1, 76) (20.4) 30.5 (12, 72) (19.4) 35 (1, 76) Month (13.4) 43.5 (34, 53) (1.4) 65 (64, 66) (12.8) 53 (34, 66) max., maximum; min., minimum. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 37

64 RESULTS Secondary outcomes For each of the foowing modes, assumptions were checked in the same way as for the primary anaysis and no significant vioations were observed. Epworth Seepiness Scae Summaries of the observed ESS score by treatment group across a time points are presented in Tabe 12. A ower score indicates a ower genera eve of daytime somnoence. There was no evidence of a difference in mean score between the treatment groups at 6 months (AMD 0.85, 95% CI 2.41 to 0.71; p = 0.28). Participants in the LTOT group were predicted to have a sighty ower ESS score at 6 months, but not statisticay significanty so. A banded (1) covariance structure was used in this mode. Numerica Rating Scae for breathessness Participants were asked to score six questions reating to their breathessness using a NRS at baseine and at 3, 6, 12, 18 and 24 months. Summaries of observed scores are presented in Tabe 13. The scores for each question were compared between the two treatment groups using a covariance pattern mode with an exchangeabe covariance pattern. TABLE 12 Observed summaries of the ESS by treatment group at each time point Unadjusted LTOT (n = 57) BMT (n = 57) Tota (n = 114) Time point n Mean (SD) Median (min., max.) n Mean (SD) Median (min., max.) n Mean (SD) Median (min., max.) Baseine (5.3) 10 (3, 24) (4.9) 10 (0, 22) (5.1) 10 (0, 24) Month (5.2) 10 (1, 24) (5.7) 10 (0, 24) (5.4) 10 (0, 24) Month (4.9) 10 (0, 18) (4.6) 11 (1, 22) (4.7) 10 (0, 22) Month (6.2) 8 (1, 23) (4.6) 9 (2, 20) (5.5) 9 (1, 23) Month (6.0) 5 (2, 21) (5.3) 10.5 (3, 21) (5.8) 7.5 (2, 21) Month (4.2) 12 (9, 15) (5.5) 9 (4, 15) max., maximum; min., minimum. TABLE 13 The NRS for breathessness score by treatment group at each time point Unadjusted LTOT (n = 57) BMT (n = 57) Tota (n = 114) Time point n Median (IQR) n Median (IQR) n Median (IQR) Q1 Baseine 57 5 (4 7) 57 5 (3 7) (4 7) Month (2 5) 53 5 (3 6) (3 6) Month (2 6) 43 5 (3 7) 88 5 (3 7) Month (4 8) 18 6 (4 7) 39 6 (4 7) Month (5 5) 8 2 (0 6) 16 5 (2 5) Month (3 6) (2 6) 38 NIHR Journas Library

65 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 TABLE 13 The NRS for breathessness score by treatment group at each time point (continued) Unadjusted LTOT (n = 57) BMT (n = 57) Tota (n = 114) Time point n Median (IQR) n Median (IQR) n Median (IQR) Q2 Baseine 57 6 (5 8) 57 5 (4 7) (4 7) Month (2 6) 53 5 (4 8) (3 7) Month (3 8) 43 6 (4 7) 88 6 (4 8) Month (4 8) 18 6 (4 7) 39 6 (4 8) Month (5 7.5) 8 3 (0 6) 16 5 (3 7) Month (1 7) (1 7) Q3 Baseine 57 3 (1 4) 57 2 (1 5) (1 5) Month (0 4) 53 2 (1 4) (1 4) Month (1 5) 43 2 (0 3) 88 2 (1 4.5) Month (1 6) 18 2 (1 4) 39 2 (1 6) Month (1 5) 8 1 (0.5 3) (1 4.5) Month (1 2) (0 2) Q4 Baseine 57 5 (2 7) 56 4 (1.5 6) (2 7) Month (0 5) 53 5 (3 6) (1 5) Month (2 7) 43 4 (1 5) 88 4 (1 6) Month (2 8) 18 5 (4 7) 39 5 (2 7) Month (0.5 5) (0 7.5) 16 4 (0 5) Month (0 4) (0 9) Q5 Baseine 57 7 (5 9) 56 7 (5 9) (5 9) Month (4 9) 53 5 (5 7) (5 9) Month (5 9) 43 5 (5 9) (5 9) Month (5 8) 18 6 (5 9) 39 7 (5 9) Month (5 10) 8 9 (7 10) (5 10) Month (8 10) (8 10) Q6 Baseine 57 7 (5 10) 57 8 (5 10) (5 10) Month (6 10) 53 7 (5 10) (5 10) Month (6 10) 43 7 (5 9) 88 8 (5 10) Month (7 9) 18 8 (5 10) 39 8 (5 10) Month (8.5 10) 8 10 (8 10) (8 10) Month (8 10) (8 10) IQR, interquartie range. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 39

66 RESULTS Q1: How bad has your breathessness fet on average over the past 24 hours? Breathessness was scored from 0 (not breathess at a) to 10 (the worst imaginabe breathessness). There was no evidence of a difference in score at 6 months (AMD 0.63, 95% CI 1.57 to 0.31; p = 0.19). Q2: What is the worst that your breathessness has been over the past 24 hours? Breathessness was scored from 0 (not breathess at a) to 10 (the worse imaginabe breathessness). There was no evidence of a difference in score at 6 months (AMD 0.16, 95% CI 1.25 to 0.94; p = 0.78). Q3: How bad is your breathessness right now? Breathessness was scored from 0 (not breathess at a) to 10 (the worse imaginabe breathessness). There was no evidence of a difference in score at 6 months (AMD 0.75, 95% CI 0.18 to 1.68; p = 0.12). Q4: How much distress has your breathessness caused you on average over the past 24 hours? Distress was scored from 0 (no distress at a) to 10 (the worse imaginabe distress). There was no evidence of a difference in score at 6 months (AMD 0.07, 95% CI 1.20 to 1.06; p = 0.90). Q5: How we have you coped with your breathessness on average over the past 24 hours? Coping was scored from 0 (I have not coped at a) to 10 (I have coped very we). There was no evidence of a difference in score at 6 months (AMD 0.79, 95% CI 0.34 to 1.93; p = 0.17). Q6: How satisfied have you fet with the treatment you have received for your breathessness? Satisfaction was scored from 0 (not satisfied at a) to 10 (competey satisfied). There was no evidence of a difference in score at 6 months (AMD 0.69, 95% CI 0.39 to 1.78; p = 0.21). Hospita Anxiety and Depression Scae The HADS was measured at baseine and at 6, 12, 18 and 24 months. The raw summary scores for the anxiety and depression subscaes are summarised in Tabes 14 and 15. The anxiety and depression subscae scores were compared between the two groups using separate covariance pattern modes both using an exchangeabe covariance pattern. No evidence of a difference was observed at 6 months in anxiety score (AMD 0.19, 95% CI 1.47 to 1.10; p = 0.77) or depression score (AMD 0.34, 95% CI 1.49 to 0.81; p = 0.56). TABLE 14 Observed summaries of the HADS anxiety subscae by treatment group at each time point Unadjusted LTOT (n = 57) BMT (n = 57) Tota (n = 114) Time point n Mean (SD) Median (min., max.) n Mean (SD) Median (min., max.) n Mean (SD) Median (min., max.) Baseine (4.6) 7 (0, 20) (4.5) 7 (0, 18) (4.5) 7 (0, 20) Month (4.5) 7 (0, 16) (4.6) 6 (0, 16) (4.6) 6 (0, 16) Month (4.2) 5 (0, 13) (3.6) 3 (0, 10) (3.9) 4 (0, 13) Month (1.6) 3.3 (0, 5) (4.6) 2 (0, 14) (3.4) 2.5 (0, 14) Month (0.7) 1.5 (1, 2) (1.0) 2 (1, 3) max., maximum; min., minimum. 40 NIHR Journas Library

67 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 TABLE 15 Observed summaries of the HADS depression subscae by treatment group at each time point Unadjusted LTOT (n = 57) BMT (n = 57) Tota (n = 114) Time point n Mean (SD) Median (min., max.) n Mean (SD) Median (min., max.) n Mean (SD) Median (min., max.) Baseine (3.5) 7 (2, 17) (3.6) 7 (1, 20) (3.5) 7 (1, 20) Month (3.3) 7 (1, 15) (3.5) 7 (1, 16) (3.4) 7 (1, 16) Month (3.2) 7 (1, 13) (4.4) 7 (1, 18) (3.8) 7 (1, 18) Month (2.2) 6.5 (3, 9) (2.5) 6.5 (1, 8) (2.3) 6.5 (1, 9) Month (5.7) 5 (1, 9) (4.0) 6 (1, 9) max., maximum; min., minimum. Karnofsky Performance Status scae of physica activity Karnofsky Performance Status is measured from 0% (death) to 100% (norma, no compaints, no signs of disease). Baseine data for the distribution of KPS scores are shown in Tabes 16 and 17. A score of 0 was given to patients for the time points after which they were known to have died. No significant difference was observed between the treatment groups at 6 months (AMD 4.97, 95% CI 2.18 to 12.13; p = 0.17). An unstructured covariance structure was used in this mode. The KPS score can be categorised as foows: 0 40 unabe to care for sef some assistance needed no specia care needed. The categorised data are shown in Tabe 18. There was no notabe difference in the distribution of scores at each time point between the tria arms. Charson Comorbidity Index The CCI was assessed at baseine and at 6, 12, 18 and 24 months (a ower score is preferabe). The observed CCI scores are summarised descriptivey by treatment group at each time point in Tabe 19. There was weak evidence of a difference between the treatment groups in CCI score at 6 months (AMD 0.45, 95% CI 0.01 to 0.91; p = 0.06), that is the BMT group were predicted as having a ower CCI score, but not statisticay significanty so. An exchangeabe covariance structure was used in this mode. The 6-minute wak test The distance a patient waked in metres over a 6-minute time period was recorded at baseine and at 6, 12, 18 and 24 months (Tabe 20). The distance waked was compared between the LTOT and BMT arms using a covariance pattern mode with an unstructured correation. No evidence of a difference in distance waked was observed between the two groups at 6 months (AMD 0.64, 95% CI to 35.83; p = 0.97). Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 41

68 RESULTS TABLE 16 Distribution of KPS scores at baseine Distribution of KPS score (%) LTOT (N = 57), n (%) BMT (N = 57), n (%) Tota (N = 114), n (%) A patients (0.0) 0 (0.0) 0 (0.0) 90 0 (0.0) 1 (1.8) 1 (0.9) 80 5 (8.8) 4 (7.0) 9 (7.9) (21.1) 11 (19.3) 23 (20.2) (49.1) 25 (43.9) 53 (46.5) (19.3) 13 (22.8) 24 (21.1) 40 1 (1.8) 3 (5.3) 4 (3.5) 30 0 (0.0) 0 (0.0) 0 (0.0) 20 0 (0.0) 0 (0.0) 0 (0.0) 10 0 (0.0) 0 (0.0) 0 (0.0) Patients in primary anaysis (0.0) 0 (0.0) 0 (0.0) 90 0 (0.0) 1 (2.0) 1 (1.0) 80 4 (7.8) 3 (5.9) 7 (6.9) (23.5) 10 (19.6) 22 (21.6) (51.0) 22 (43.1) 48 (47.1) 50 9 (17.7) 12 (23.5) 21 (20.6) 40 0 (0.0) 3 (5.9) 3 (2.9) 30 0 (0.0) 0 (0.0) 0 (0.0) 20 0 (0.0) 0 (0.0) 0 (0.0) 10 0 (0.0) 0 (0.0) 0 (0.0) TABLE 17 Distribution of KPS scores at baseine for patients recruited by 1 May 2013 Distribution of KPS score (%) LTOT (N = 25), n (%) NOT (N = 25), n (%) BMT (N = 24), n (%) Tota (N = 114), n (%) (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 90 0 (0.0) 0 (0.0) 1 (4.2) 1 (1.4) 80 2 (8.0) 1 (4.0) 2 (8.3) 5 (6.8) 70 7 (28.0) 4 (16.0) 4 (16.7) 15 (20.3) (48.0) 10 (40.0) 10 (41.7) 32 (43.2) 50 3 (12.0) 8 (32.0) 5 (20.8) 16 (21.6) 40 1 (4.0) 1 (4.0) 2 (8.3) 4 (5.4) 30 0 (0.0) 1 (4.0) 0 (0.0) 1 (1.4) 20 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 10 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 42 NIHR Journas Library

69 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 TABLE 18 Karnofsky Performance Status scae summaries by treatment group at each time point KPS summary LTOT (n = 57) BMT (n = 57) Tota (n = 114) Baseine Dead, n (%) 0 (0.0) 0 (0.0) 0 (0.0) Unabe to care for sef, n (%) 1 (1.8) 3 (5.3) 4 (3.5) Some assistance needed, n (%) 51 (89.5) 49 (86.0) 100 (87.7) No specia care needed, n (%) 5 (8.8) 5 (8.8) 10 (8.8) Median (min., max.) 60 (40, 80) 60 (40, 90) 60 (40, 90) Month 3 Dead, n (%) 3 (5.4) 3 (5.4) 6 (5.4) Unabe to care for sef, n (%) 2 (3.6) 5 (8.9) 7 (6.3) Some assistance needed, n (%) 46 (82.1) 43 (76.8) 89 (79.5) No specia care needed, n (%) 5 (8.9) 5 (8.9) 10 (8.9) Median (min., max.) 60 (0, 90) 60 (0, 80) 60 (0, 90) Month 6 Dead, n (%) 4 (8.2) 6 (12.5) 10 (10.3) Unabe to care for sef, n (%) 2 (4.1) 3 (6.3) 5 (5.2) Some assistance needed, n (%) 36 (73.5) 35 (72.9) 71 (73.2) No specia care needed, n (%) 7 (14.3) 4 (8.3) 11 (11.3) Median (min., max.) 60 (0, 80) 60 (0, 80) 60 (0, 80) Month 12 Dead, n (%) 6 (22.2) 11 (37.9) 17 (30.4) Unabe to care for sef, n (%) 0 (0.0) 0 (0.0) 0 (0.0) Some assistance needed, n (%) 18 (66.7) 17 (58.6) 35 (62.5) No specia care needed, n (%) 3 (11.1) 1 (3.5) 4 (7.1) Median (min., max.) 60 (0, 80) 60 (0, 80) 60 (0, 80) Month 18 Dead, n (%) 6 (40.0) 12 (60.0) 18 (51.4) Unabe to care for sef, n (%) 0 (0.0) 0 (0.0) 0 (0.0) Some assistance needed, n (%) 9 (60.0) 7 (35.0) 16 (45.7) No specia care needed, n (%) 0 (0.0) 1 (5.0) 1 (2.9) Median (min., max.) 60 (0, 70) 0 (0, 80) 0 (0, 80) Month 24 Dead, n (%) 6 (75.0) 12 (92.3) 18 (85.7) Unabe to care for sef, n (%) 0 (0.0) 0 (0.0) 0 (0.0) Some assistance needed, n (%) 1 (12.5) 1 (7.7) 2 (9.5) No specia care needed, n (%) 1 (12.5) 0 (0.0) 1 (4.8) Median (min., max.) 0 (0, 80) 0 (0, 60) 0 (0, 80) max., maximum; min., minimum. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 43

70 RESULTS TABLE 19 Charson Comorbidity Index scores by treatment group at each time point Unadjusted LTOT (n = 57) BMT (n = 57) Tota (n = 114) Time point n Mean (SD) Median (min., max.) n Mean (SD) Median (min., max.) n Mean (SD) Median (min., max.) Baseine (2.0) 6 (3, 11) (2.3) 7 (1, 12) (2.1) 7 (1, 12) Month (2.2) 7 (3, 15) (2.1) 7 (2, 12) (2.1) 7 (2, 15) Month (1.9) 6 (3, 11) (1.7) 8 (5, 11) (1.9) 7 (3, 11) Month (2.3) 6 (4, 11) (1.8) 6 (6, 11) (2.0) 6 (4, 11) Month (2.1) 5.5 (4, 7) (1.5) 5 (4, 7) max., maximum; min., minimum. TABLE 20 Six-minute wak test distance by treatment group at each time point Unadjusted LTOT (n = 57) BMT (n = 57) Tota (n = 114) Time point n Median (IQR) n Median (IQR) n Median (IQR) Baseine (54 210) ( ) (54 210) Month (60 240) (60 250) (60 248) Month (90 230) (50 245) (70 245) Month ( ) (70 340) ( ) Month IQR, interquartie range. Prevaence of hypoxia The arteria oxygen saturation measured as part of the 6MWT protoco (i.e. at rest, during peak exercise and then during recovery from exercise) is shown in Tabe 21. The prevaences of oxygen saturation in the ranges < 90%, 90% to < 95% and 95% at rest, at peak and 5 minutes after the 6MWT are summarised Tabes 22, 23 and 24, respectivey. N-termina pro-b-type natriuretic hormone Patients NT-proBNP eve was measured at baseine and at 3, 6, 12, 18 and 24 months. The NT-proBNP eves are summarised descriptivey by treatment group at each time point in Tabe 25. There was a wide range of NT-proBNP eves at a time points; for exampe, at baseine, the range was 82 35,000. NT-proBNP eve was compared between the two treatment groups at 6 months with a covariance pattern mode, using an unstructured correation, adjusting for baseine eve. NT-proBNP eve data were highy positivey skewed at a time points and so were og transformed. There was no evidence of a difference between the treatment groups at 6 months (p = 0.80). As the outcome was og transformed, interpretation of the estimated difference between the two treatment groups is a itte more difficut. The LTOT group was predicted to have a decrease in og-nt-probnp of 0.04 (95% CI 0.31 to 0.24) or, in other words, patients in the LTOT group were expected to have a NT-proBNP eve 0.96 times that of a BMT patient (95% CI 0.73 to 1.27). 44 NIHR Journas Library

71 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 TABLE 21 Arteria oxygen saturation in the 6MWT LTOT (n = 57) BMT (n = 57) Tota (n = 114) Time point n Median (min., max.) n Median (min., max.) n Median (min., max.) SaO 2 at rest (%) Baseine (91, 99) (92, 99) (91, 99) Month (92, 99) (91, 99) (91, 99) Month (93, 100) (95, 100) (93, 100) Month (93, 98) 6 96 (95, 98) (93, 98) Month SaO 2 at peak (%) Baseine (90, 100) (89, 100) (89, 100) Month (78, 100) (89, 100) (78, 100) Month (93, 100) (93, 100) (93, 100) Month (94, 100) 6 97 (96, 99) (94, 100) Month SaO 2 post test (%) Baseine (91, 100) (91, 100) (91, 100) Month (92, 100) (92, 100) (92, 100) Month (89, 100) (95, 100) (89, 100) Month (94, 98) 6 96 (94, 99) (94, 99) Month max., maximum; min., minimum; SaO 2, arteria oxygen saturation. TABLE 22 Oxygen saturation measured before the 6MWT, by treatment group at each time point Unadjusted Oxygen saturation before 6MWT LTOT (n = 57) BMT (n = 57) Tota (n = 114) < 90%, n (%) 90% to < 95%, n (%) 95%, n (%) < 90%, n (%) 90% to < 95%, n (%) 95%, n (%) < 90%, n (%) 90% to < 95%, n (%) 95%, n (%) Baseine 0 (0.0) 10 (18.2) 45 (81.8) 0 (0.0) 3 (5.3) 54 (94.7) 0 (0.0) 13 (11.6) 99 (88.4) Month 6 0 (0.0) 13 (31.7) 28 (68.3) 0 (0.0) 2 (6.1) 31 (93.9) 0 (0.0) 15 (20.3) 59 (79.7) Month 12 0 (0.0) 2 (10.5) 17 (89.5) 0 (0.0) 0 (0.0) 14 (100.0) 0 (0.0) 2 (6.1) 31 (93.9) Month 18 0 (0.0) 1 (16.7) 5 (83.3) 0 (0.0) 0 (0.0) 6 (100.0) 0 (0.0) 1 (8.3) 11 (91.7) Month 24 0 (0.0) 0 (0.0) 1 (100.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 1 (100.0) Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 45

72 RESULTS TABLE 23 Oxygen saturation measured at peak (maximum saturation observed during 6MWT), by treatment group at each time point Unadjusted LTOT (n = 57) BMT (n = 57) Tota (n = 114) Time point < 90%, n (%) 90% to < 95%, n (%) 95%, n (%) < 90%, n (%) 90% to < 95%, n (%) 95%, n (%) < 90%, n (%) 90% to < 95%, n (%) 95%, n (%) Baseine 0 (0.0) 12 (23.5) 39 (76.5) 1 (1.9) 6 (11.5) 45 (86.5) 1 (1.0) 18 (17.5) 39 (76.5) Month 6 2 (5.3) 8 (21.1) 28 (73.7) 1 (3.1) 3 (9.4) 28 (87.5) 3 (4.3) 11 (15.7) 56 (80.0) Month 12 0 (0.0) 4 (22.2) 14 (77.8) 0 (0.0) 1 (7.1) 13 (92.9) 0 (0.0) 5 (15.6) 27 (84.4) Month 18 0 (0.0) 1 (16.7) 5 (83.3) 0 (0.0) 0 (0.0) 6 (100.0) 0 (0.0) 1 (8.3) 11 (91.7) Month 24 0 (0.0) 0 (0.0) 1 (100.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 1 (100.0) TABLE 24 Oxygen saturation measured at 5 minutes post 6MWT, by treatment group at each time point Unadjusted LTOT (n = 57) BMT (n = 57) Tota (n = 114) Time point < 90%, n (%) 90% to < 95%, n (%) 95%, n (%) < 90%, n (%) 90% to < 95%, n (%) 95%, n (%) < 90%, n (%) 90% to < 95%, n (%) 95%, n (%) Baseine 0 (0.0) 7 (12.7) 48 (87.3) 0 (0.0) 7 (12.5) 49 (87.5) 0 (0.0) 14 (12.6) 97 (87.4) Month 6 0 (0.0) 9 (22.5) 31 (77.5) 0 (0.0) 3 (9.1) 30 (90.9) 0 (0.0) 12 (16.4) 61 (83.6) Month 12 1 (5.6) 1 (5.6) 16 (88.9) 0 (0.0) 0 (0.0) 14 (100.0) 1 (3.1) 1 (3.1) 30 (93.8) Month 18 0 (0.0) 1 (16.7) 5 (83.3) 0 (0.0) 1 (16.7) 5 (83.3) 0 (0.0) 2 (16.7) 10 (83.3) Month 24 0 (0.0) 0 (0.0) 1 (100.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 1 (100.0) TABLE 25 Leve of NT-proBNP by treatment group at each time point Unadjusted LTOT (n = 57) BMT (n = 57) Tota (n = 114) Time point n Mean (SD) Median (min., max.) n Mean (SD) Median (min., max.) n Mean (SD) Median (min., max.) Baseine (8402.0) 2243 (118, 35,000) (4026.7) 1931 (82, 15,594) (6627.3) (82, 35,000) Month (6965.9) 1713 (138, 35,000) (5714.9) 1984 (25.4, 29,331) (6340.0) (25.4, 35,000) Month (7507.4) 1916 (94, 35,000) (5401.1) 1554 (38.3, 32,621) (6561.9) 1608 (38.3, 35,000) Month (7358.9) 1622 (91, 35,000) (3297.3) 1803 (140.4, 14,208) (5889.3) (91, 35,000) Month (1559.1) (264, 4639) (31,78.5) 1650 (222, 9367) (2393.5) 1650 (222, 9367) Month max., maximum; min., minimum. 46 NIHR Journas Library

73 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Left ventricuar ejection fraction Patients had an echocardiogram at baseine and at 12 and 24 months. Where possibe, the LVEF was cacuated and expressed as a percentage. To be eigibe for the HOT tria, participants had to have a LVEF < 40% or have at east moderate LV impairment. LVEF and LV impairment at the three time points are summarised in Tabe 26. One participant had a LVEF above 40% at baseine; however, this participant had LV impairment graded severe and was thus sti eigibe for the tria. Ony one patient at 12 months had a LVEF about 40%. At 24 months, echocardiography findings were avaiabe for ony two participants, both in the LTOT group. These two patients were graded as having severe LV impairment, but the LVEF coud be cacuated for ony one patient (a vaue of 25). Mortaity At 6 months, 15 patients had died (LTOT n = 4, NOT n = 5 and BMT n = 6). A further 10 deaths were reported after 6 months of foow-up (LTOT n = 2, NOT n = 2 and BMT n = 6). Mortaity was anaysed as a time-to-event outcome. The main anaysis compared the BMT and LTOT treatment groups. For each group, the distribution of time from randomisation to death is described using Kapan Meier surviva curves (Figure 8). Unadjusted Cox regression gave a HR of 2.03 (95% CI 0.76 to 5.40) for BMT reative to LTOT, but this was not statisticay significant (p = 0.16). Adjusting for baseine CCI score, the HR was sighty ower (1.84, 95% CI 0.68 to 4.96; p = 0.30), indicating, again, that the risk of death was higher in the BMT group than in the LTOT group, but the difference was not statisticay significant (p = 0.30). The CCI was not a significant predictor in the surviva mode (p = 0.19) Additiona surviva anayses were undertaken comparing the NOT and LTOT arms ony incuding patients randomised up to the time that the NOT arm was dropped from the study. The number of patients recruited to the LTOT, NOT and BMT arms was 25, 25 and 24, respectivey (recruited up to the end of Apri 2013). There was no significant difference in surviva between the groups (Kapan Meier, χ 2 = 2.07, df = 1; p = 0.15). The oxygen arms were combined and compared against the BMT arm incuding ony those patients randomised up to the time that the NOT arm was dropped. Figure 9 shows the surviva curve for BMT against LTOT and NOT. There was no significant difference in surviva between the groups (Kapan Meier, χ 2 = 1.12, df = 1; p = 0.29). The 75th percentie for the BMT group was days (SE 79.3 days) and for the combined LTOT and NOT group it was days (SE 0.98 days). Unadjusted Cox regression gave a HR of 1.64 (95% CI 0.65 to 4.17), indicating that the risk of death was higher in the BMT group than in the LTOT and NOT groups, but the difference was not statisticay significant (p = 0.30). The mode adjusting for baseine CCI score gave very simiar estimates and CCI was not a significant predictor of mortaity. The proportiona hazards assumption was checked using og og pots of the estimated survivor function, potting scaed Schoenfed residuas and by the Grambsch and Therneau test. Visua inspection of the og og and Schoenfed pots indicated potentia non-proportionaity of the treatment group variabe in each mode; however, the Grambsch and Therneau test did not indicate evidence of non-proportionaity. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 47

74 RESULTS TABLE 26 Left ventricuar ejection fraction and severity of LV dysfunction by treatment group at each time point Unadjusted Time point LTOT (n = 57) BMT (n = 57) Tota (n = 114) Baseine LVEF (%) n Mean (SD) 28.0 (7.7) 28.2 (8.1) 28.1 (7.9) Median (min., max.) 29 (7, 39) 28 (11, 50) 28 (7, 50) Severity of LV dysfunction, n (%) Mid 0 (0.0) 0 (0.0) 0 (0.0) Mid to moderate 1 (1.8) 0 (0.0) 1 (0.9) Moderate 11 (19.3) 9 (15.8) 20 (17.5) Moderate to severe 10 (17.5) 12 (21.1) 22 (19.3) Severe 35 (61.4) 36 (63.2) 71 (62.3) Month 12 LVEF (%) n Mean (SD) 30.5 (7.5) 25.2 (9.1) 28.3 (8.5) Median (min., max.) 30 (17, 47.5) 23 (8, 38) 30 (8, 47.5) Severity of LV dysfunction, n (%) Mid 0 (0.0) 1 (6.3) 1 (2.8) Mid to moderate 1 (5.0) 0 (0.0) 1 (2.8) Moderate 4 (20.0) 2 (12.5) 6 (16.7) Moderate to severe 3 (15.0) 1 (6.3) 4 (11.1) Severe 12 (60.0) 12 (75.0) 24 (66.7) Month 24 LVEF (%) n Mean (SD) Median (min., max.) Severity of LV dysfunction, n (%) Mid 0 (0.0) 0 (0.0) Mid to moderate 0 (0.0) 0 (0.0) Moderate 0 (0.0) 0 (0.0) Moderate to severe 0 (0.0) 0 (0.0) Severe 2 (100.0) 2 (100.0) max., maximum; min., minimum. 48 NIHR Journas Library

75 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO Cumuative surviva Group BMT LTOT BMT-censored LTOT-censored FIGURE 8 Kapan Meier surviva curve. Surviva Cumuative surviva Group BMT LTOT and NOT BMT (censored) LTOT and NOT (censored) Surviva (days) FIGURE 9 Kapan Meier surviva curve for BMT vs. LTOT pus NOT. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 49

76 RESULTS Number of days aive and out of hospita The number of DAOH was cacuated for each patient. Summaries of the number and percentage of days the patients in the groups were aive and out of hospita are presented in Tabe 27. European Quaity of Life-5 Dimensions The primary objective of the HOT tria was to assess the HRQoL benefits of HOT in the management of patients with stabe CHF who are sti severey symptomatic despite maximay toerated medica therapy. A preference-based measure of heath status is the EQ-5D instrument (EQ-5D-3L). The EQ-5D is a descriptive system that aows patients to indicate their current heath state across five dimensions: mobiity, sef-care, usua activities, pain/discomfort and anxiety/depression. In the EQ-5D-3L questionnaire provided to patients in the HOT study, each dimension has three eves: no probems, some probems or extreme probems. Patients seect a eve for each of the five dimensions in order to generate a heath state description, which can then be converted into a singe summary score. The summary score is anchored at 1, which indicates fu heath, and 0, which is equivaent to death. States worse than death are possibe. Patients were asked to compete the EQ-5D-3L at baseine and at 6, 12, 18 and 24 months. The EQ-5D-3L heath state descriptions coected in the HOT tria were converted to summary scores by appying a formua based on vaues coected in the UK genera popuation. 69 Tabe 28 shows the summary EQ-5D-3L scores in the HOT tria overa and by tria arm at each time point. On average, participants in the HOT tria had a ow HRQoL. A mean score of 0.52 at baseine indicates that patients woud exchange 10 years of ife in their current heath state for 5.2 years in fu heath and, thus, are wiing to sacrifice 4.8 years of ife. An unadjusted t-test comparison of overa baseine scores to each time point suggested that there was a statistica significant increase in mean HRQoL (p < 0.05) between 18 months and baseine, and between 24 months and baseine. However, by 18 months foow-up few patients provided data for the EQ-5D (n = 5). A simpe anaysis of variance comparing each treatment at each time point suggested that there was no evidence of a difference in scores between treatment arms (p > 0.05). An unadjusted t-test comparison of baseine scores to each time point for each treatment showed ony one statisticay significant resut (p < 0.05): for patients aocated to NOT at 6 months compared with baseine. However, the tota number of patients in the NOT arm was sma (n = 25), and ony two-thirds (n = 17) competed the EQ-5D-3L questionnaire at 6 months. TABLE 27 Days aive and out of hospita BMT LTOT NOT Tota Number of DAOH Number of days % Number of days % Number of days % Number of days % Mean (SD) (171.1) 85.4 (27.7) (182.0) 89.8 (25.6) (236.6) 74.3 (38.2) (190.8) 85.2 (29.3) Min., max. 21, , , 752 4, 100 3, 730 1, 100 3, 773 1, 100 max., maximum; min., minimum. 50 NIHR Journas Library

77 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 TABLE 28 Summary of EQ-5D-3L scores for each treatment group by time point Unadjusted LTOT (n = 57) NOT (n = 25) BMT (n = 57) Tota (n = 114) Time point n Mean (SD) Min., max. n Mean (SD) Min., max. n Mean (SD) Min., max. n Mean (SD) Min., max. Baseine (0.24) 0.09, (0.32) 0.24, (0.27) 0.18, (0.26) 0.24, 0.88 Month (0.23) 0.00, (0.28) 0.02, (0.30) 0.07, (0.27) 0.07, 1 Month (0.36) 0.18, (0.25) 0.10, (0.25) 0.07, (0.30) 0.18, 1 Month (0.30) 0.13, (0.34) 0.06, (0.12) 0.52, (0.27) 0.06, 1 Month (0.02) 0.81, (0.36) 0.00, (0.35) 0.00, 0.85 max., maximum; min., minimum. Adverse events The foowing sections detai the SAEs and non-saes for a three groups. Serious adverse events In tota, 67 participants had 123 SAEs. More participants in the BMT arm than in the LTOT and NOT arms experienced one or more SAEs (52.6% compared with 45.6% and 44.0% respectivey). However, these differences were not statisticay significant (χ 2 = 0.78, df = 2; p = 0.68). Of these SAEs, 1.6% were cassified as ife-threatening and 19.5% were deaths. In tota, there were 25 deaths but a SAE form was not submitted for one of these; instead, it was recorded as the outcome of a SAE for hospitaisation. The majority of the deaths were not deemed to be reated to treatment (80.5%) and 41.5% were expected. Detais of a SAEs reported are shown in Tabe 29. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 51

78 RESULTS TABLE 29 Serious adverse events Events LTOT (n = 57) NOT (n = 25) BMT (n = 57) Tota (n = 139) Number of participants with one or more adverse events, n (%) 26 (45.6) 11 (44.0) 30 (52.6) 67 (48.2) Tota number of adverse events Events per participant, n (%) 1 17 (65.4) 5 (45.5) 19 (63.3) 41 (61.2) 2 3 (11.5) 3 (27.3) 5 (16.7) 11 (16.4) 3 3 (11.5) 2 (18.2) 2 (6.7) 7 (10.4) 4 3 (11.5) 0 (0.0) 1 (3.3) 4 (6.0) 5 0 (0.0) 0 (0.0) 2 (6.7) 2 (3.0) 6 0 (0.0) 1 (9.0) 0 (0.0) 1 (1.5) 7 0 (0.0) 0 (0.0) 1 (3.3) 1 (1.5) Event detais (a), n (%) Death 5 (11.4) 7 (30.4) 12 (21.4) 24 (19.5) Hospita proonged 2 (4.6) 1 (4.4) 2 (3.6) 5 (4.1) Hospitaisation 35 (79.5) 15 (65.2) 41 (73.2) 91 (74.0) Life-threatening 1 (2.3) 0 (0.0) 1 (1.8) 2 (1.6) Other 1 (2.3) 0 (0.0) 0 (0.0) 1 (0.8) Outcome, n (%) Died 6 (13.6) 7 (30.4) 12 (21.4) 25 (20.3) Ongoing 10 (22.7) 3 (13.0) 18 (32.1) 31 (25.2) Recovered fuy 24 (54.5) 10 (43.5) 25 (44.6) 59 (48.0) Recovered partiay 4 (9.1) 3 (13.0) 1 (1.8) 8 (6.5) Reationship to treatment, n (%) Not abe to assess 0 (0.0) 0 (0.0) 1 (1.8) 1 (0.8) Not reated 32 (72.7) 21 (91.3) 46 (82.1) 99 (80.5) Unikey reated 12 (27.3) 2 (8.7) 9 (16.1) 23 (18.7) Expected or unexpected, n (%) Expected 19 (43.2) 11 (47.8) 21 (37.5) 51 (41.5) Unexpected 25 (56.8) 12 (52.2) 35 (62.5) 72 (58.5) 52 NIHR Journas Library

79 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Foow-up for serious adverse events In tota, 17 participants had foow-up for their severe adverse event. This is summarised in Tabe 30. Non-serious adverse events In tota, 32 participants had 139 non-saes. More participants in the BMT and LTOT arms than in the NOT arm experienced one or more non-saes (26.3%, 26.3% and 8.0% respectivey). Of these adverse events, 72.7% were cassified as mid, 50.9% resoved within the course of the tria, a were deemed not, or unikey to be, reated to treatment and 17.5% were expected. Detais of a adverse events reported are shown in Tabe 31. TABLE 30 Foow-up SAEs Events LTOT (n = 57) NOT (n = 25) BMT (n = 57) Tota (n = 139) Number of participants with one or more foow-ups Tota number of events Events per participant, n (%) 1 5 (71.4) 1 (100.0) 7 (77.8) 13 (76.4) 2 2 (28.6) 0 (0.0) 0 (0.0) 2 (11.8) 3 0 (0.0) 0 (0.0) 2 (22.2) 2 (11.8) Intensity, n (%) Mid 4 (44.4) 0 (0.0) 1 (7.7) 5 (21.7) Moderate 0 (0.0) 0 (0.0) 8 (61.5) 8 (34.8) Severe 5 (55.6) 1 (100.0) 4 (30.8) 10 (43.5) Outcome, n (%) Ongoing 3 (33.3) 0 (0.0) 5 (38.5) 8 (34.8) Resoved 6 (66.7) 1 (100.0) 8 (61.5) 15 (65.2) Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 53

80 RESULTS TABLE 31 Non-SAEs Non-SAEs LTOT (n = 57) NOT (n = 25) BMT (n = 57) Tota (n = 139) Number of participants with one or more adverse events, n (%) 16 (28.1) 2 (8.0) 15 (26.3) 33 (23.7) Tota number of adverse events Events per participant, n (%) 1 9 (52.9) 1 (50.0) 11 (73.3) 21 (63.6) 2 3 (18.8) 0 (0.0) 3 (20.0) 6 (18.2) 3 1 (6.3) 0 (0.0) 1 (6.7) 2 (6.1) 4 2 (12.5) 0 (0.0) 0 (0.0) 2 (6.1) 5 1 (6.3) 1 (50.0) 0 (0.0) 2 (6.1) Intensity, n (%) Missing 0 (0.0) 0 (0.0) 1 (5.0) 1 (1.8) Mid 24 (77.4) 6 (100.0) 12 (60.0) 42 (73.7) Moderate 7 (22.6) 0 (0.0) 6 (30.0) 13 (22.8) Severe 0 (0.0) 0 (0.0) 1 (5.0) 1 (1.8) Outcome, n (%) Ongoing 12 (38.7) 0 (0.0) 7 (35.0) 19 (33.3) Ongoing with sequeae 2 (6.5) 0 (0.0) 2 (10.0) 4 (7.0) Resoved 16 (51.6) 4 (66.7) 9 (45.0) 29 (50.9) Resoved with sequeae 1 (3.2) 2 (33.3) 2 (10.0) 5 (8.8) Reationship to treatment, n (%) Not reated 27 (87.1) 6 (100.0) 20 (100.0) 53 (93.0) Unikey reated 4 (12.9) 0 (0.0) 0 (0.0) 4 (7.0) Expected or unexpected, n (%) Expected 4 (12.9) 0 (0.0) 6 (30.0) 10 (17.5) Unexpected 27 (87.1) 6 (100.0) 14 (70.0) 47 (82.5) Adherence The oxygen concentrators were generay instaed in the homes of patients aocated to receive HOT within 2 weeks of randomisation (median 5 days). Two participants did not receive a machine as one died and one withdrew from the tria very shorty after randomisation. Eight (14%) patients in the LTOT arm formay withdrew from the tria treatment throughout foow-up and requested that the oxygen concentrator be removed from their home. Here, we present data on patient-reported adherence and data from the oxygen suppiers meter readings. 54 NIHR Journas Library

81 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Patient-reported adherence Patients aocated to LTOT were asked to use their home oxygen concentrators for at east 15 hours per day incuding overnight hours, and patients in the NOT arm to use it overnight (8 hours). At 3, 6, 12, 18 and 24 months, patients aocated to either of the HOT arms were asked the foowing question: Thinking about the past month, typicay for how many hours a day did you use your HOT? Of the patients who provided a response to this question in the LTOT arm, the proportion reporting that they did not use the oxygen therapy increased at each subsequent time point (Tabe 32). Among patients who reported that they used the oxygen, the majority used it for ess than the recommended 15 hours. In the NOT arm, between 44% and 100% of patients reported using the oxygen for the recommended duration or more (at east 8 hours) at each time point. TABLE 32 Patient-reported adherence to HOT Adherence LTOT (n = 57) NOT (n = 25) Month 3, n n= 54 n = 20 None, I do not use it 8 (14.8) 2 (10.0) Less than 8 hours a day 22 (40.7) 6 (30.0) Overnight but ess than 15 hours a day 19 (35.2) 10 (50.0) 15 hours a day 5 (9.3) 2 (10.0) Month 6, n n= 44 n = 17 None, I do not use it 9 (20.5) 3 (17.7) Less than 8 hours a day 16 (36.4) 5 (29.4) Overnight but ess than 15 hours a day 14 (31.8) 8 (47.1) 15 hours a day 5 (11.4) 1 (5.9) Month 12, n n= 22 n = 17 None, I do not use it 6 (27.3) 5 (29.4) Less than 8 hours a day 9 (40.9) 3 (17.7) Overnight but ess than 15 hours a day 4 (18.2) 8 (47.1) 15 hours a day 3 (13.6) 1 (5.9) Month 18, n n= 9 n = 9 None, I do not use it 5 (55.6) 2 (22.2) Less than 8 hours a day 1 (11.1) 3 (33.3) Overnight but ess than 15 hours a day 2 (22.2) 4 (44.4) 15 hours a day 1 (1.1) 0 (0.0) Month 24, n n= 2 n = 2 None, I do not use it 2 (100.0) 0 (0.0) Less than 8 hours a day 0 (0.0) 0 (0.0) Overnight but ess than 15 hours a day 0 (0.0) 0 (0.0) 15 hours a day 0 (0.0) 2 (100.0) Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 55

82 RESULTS Oxygen suppiers meter readings The oxygen concentrators instaed in the homes of the participants aocated to the LTOT and NOT arms contained meters which recorded the number of hours the machine was used for. Meter readings were taken when the concentrators were instaed and again at intermittent intervas (range of days between visits, 8 469). From these data, an average daiy usage of the machines in hours per participant coud be cacuated. Data were suppied for 57 participants in the LTOT arm and for 24 patients in the NOT arm. In the case of four patients, no visit was recorded after the initia instaation and so average daiy usage coud not be cacuated. It was hoped that the oxygen concentrators woud be instaed in the participant s home as quicky as possibe after they were randomised. One patient aready had a machine in their home that they had previousy been using and so continued to use this machine during the tria foow-up. Two participants did not receive a machine, as one died and one withdrew from the tria very shorty after randomisation. Of the other patients, the median time between randomisation and instaation of the machine was 5 days. The vast majority of patients received the machine between 0 and 16 days after randomisation, one received it after 35 days and another after 91 days (owing to a suppy issue at the Dundee site). It was recorded that one patient ony received their machine 420 days after randomisation. This was queried with the oxygen suppier (Air Products) who suggested the patient may have ived at another address at randomisation and ony the records reating to the second address were sent to us for anaysis. For those patients who had a second visit (n = 77 patients), the machine-use data are presented in Tabe 33, which shows the number of patients (%) who, on average, per day, used the machine in the foowing categories: < 8hours,8 15 hours and >15 hours. Summary statistics of the average daiy use are presented, for each oxygen therapy group and overa. The figures are based on average houry use between first and ast meter readings. Average usage between readings fuctuated, indicating that patients might not have used the machine consistenty throughout the foow-up. In both arms, the majority of participants were using the machine for, on average, ess than 8 hours per day. The patients in the NOT arm, despite being asked to use the oxygen machine for ony 8 hours (overnight), tended to use the machine for onger than those aocated to LTOT, who were asked to use the machine for 15 hours per day (incuding overnight) (median daiy usage 4.9 hours compared with 3.8 hours). No forma statistica tests were conducted on these data and figures are for information ony. TABLE 33 Oxygen machine usage data Oxygen machine usage data LTOT (n = 55) NOT (n = 23) Overa (n = 77) Average hours of use per day < 8, n (%) 37 (67.3) 19 (82.6) 56 (71.8) 8 15, n (%) 17 (30.9) 3 (13.0) 20 (25.6) > 15, n (%) 1 (1.8) 1 (4.4) 2 (2.6) Tota hours Mean (SD) 5.4 (5.0) 5.4 (4.4) 5.4 (4.8) Median (min., max.) 3.8 (0.1, 15.5) 4.9 (0, 19.6) 4.2 (0, 19.6) max., maximum; min., minimum. 56 NIHR Journas Library

83 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Overnight Embetta seep study If an Embetta device was ocay accessibe, participants were asked to compete an overnight seep test at baseine and at 6, 12, 18 and 24 months. The Embetta records mutipe signas to detect whether or not a patient has seep apnoea, either obstructive or centra. A pair of nasa prongs records air fow and a puse oximeter worn on the finger records arteria oxygen saturation. Together, these signas detect episodes of reduced breathing (hypopnoea) or arrested breathing (apnoea). A band is worn across the chest and upper abdomen: this channe records respiratory effort. If decreased nasa air fow (and decrease in arteria oxygen saturation) suggests apnoea or hypopnoea, and if respiratory efforts are detected, an obstructive event is recorded. If, on the other hand, no respiratory efforts are detected, a centra event is recorded. The apnoea hypopnoea index (AHI) is used to indicate the severity of seep apnoea and represents the number of apnoea and hypopnoea events per hour of seep. AHI vaues are categorised as norma (0 4), mid seep apnoea (5 14), moderate seep apnoea (15 29) and severe seep apnoea (30+). Considering that both AHI and oxygen saturation can give an overa assessment of the seep apnoea severity. The AHI, the number of desaturations per hour and the tota seep time with saturation ess than 90% at each time point are summarised by treatment group in Tabe 34. The seep test was conducted in ony five sites (Barnet, Darington, Dundee, Durham and Hu), and ony if the patient consented to the test, hence the ow number of tests conducted. There was a variety of potentia probems with the seep study. Some patients found the Embetta difficut to set up and use despite training at the hospita; some found that it interfered with seep and were unwiing to use it again in the event of a poor-quaity recording. As there is no currenty recognised indication for seep studies in heart faiure patients per se, it is ikey that some investigators did not encourage patients to take part in this substudy. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 57

84 RESULTS TABLE 34 Resuts from the Embetta seep test by treatment group at each time point Unadjusted Time point LTOT (n = 57) BMT (n = 57) Tota (n = 114) Baseine n AHI, n (%) Norma 15 (46.9) 20 (52.6) 35 (50.0) Mid 8 (25.0) 11 (29.0) 19 (27.1) Moderate 6 (18.8) 6 (15.8) 12 (17.1) Severe 3 (9.4) 1 (2.6) 4 (5.7) Desaturations per hour Mean (SD) 8.9 (10.6) 10.6 (12.4) 9.8 (11.6) Median (min., max.) 5.1 (0, 43.6) 5.7 (0, 59.6) 5.4 (0, 59.6) Seep time with SaO 2 < 90% (minutes) Mean (SD) 0.03 (0.1) 0.03 (0.1) 0.03 (0.1) Median (min., max.) 0 (0, 0.8) 0 (0, 0.6) 0 (0, 0.8) Month 6 n AHI, n (%) Norma 6 (35.3) 9 (39.1) 15 (37.5) Mid 7 (41.2) 8 (34.8) 15 (37.5) Moderate 3 (17.7) 5 (21.7) 8 (20.0) Severe 1 (5.9) 1 (4.4) 2 (5.0) Desaturations per hour Mean (SD) 9.3 (11.4) 10.5 (11.0) 10.0 (11.0) Median (min., max.) 4 (0, 36.4) 7.6 (0, 37) 5.2 (0, 37) Seep time with SaO 2 < 90% (minutes) Mean (SD) 0 (0.0) 0.02 (0.1) 0.02 (0.1) Median (min., max.) 0 (0, 0) 0 (0, 0.6) 0 (0, 0.6) Month 12 n AHI, n (%) Norma 7 (70.0) 4 (50.0) 11 (61.1) Mid 2 (20.0) 1 (12.5) 3 (16.7) Moderate 1 (10.0) 3 (37.5) 4 (22.2) Severe 0 (0.0) 0 (0.0) 0 (0.0) Desaturations per hour Mean (SD) 4.4 (3.5) 11.7 (11.4) 7.6 (8.6) Median (min., max.) 3.2 (0.4, 11) 7.0 (0.4, 27.7) 3.8 (0.4, 27.7) 58 NIHR Journas Library

85 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 TABLE 34 Resuts from the Embetta seep test by treatment group at each time point (continued) Unadjusted Time point LTOT (n = 57) BMT (n = 57) Tota (n = 114) Seep time with SaO 2 < 90% (minutes) Mean (SD) 0 (0.0) 0.01 (0.03) 0.01 (0.02) Median (min., max.) 0 (0, 0) 0 (0, 0.1) 0 (0, 0.1) Month 18 n AHI, n (%) Norma 2 (40.0) 1 (33.3) 3 (37.5) Mid 0 1 (33.3) 1 (12.5) Moderate 2 (40.0) 1 (33.3) 3 (37.5) Severe 1 (20.0) 0 (0.0) 1 (12.5) Desaturations per hour Mean (SD) 8.3 (7.8) 7.7 (11.3) 8.0 (8.5) Median (min., max.) 8.2 (0.2, 16.4) 2.2 (0.2, 20.7) 3.2 (0.2, 20.7) Seep time with SaO 2 < 90% (minutes) Mean (SD) 0 (0.0) 0.03 (0.06) 0.01 (0.04) Median (min., max.) 0 (0, 0) 0 (0, 0.1) 0 (0, 0.1) Month 24 n AHI, n (%) Norma 1 (100.0) 0 (0.0) 0 (0.0) Mid 0 (0.0) 0 (0.0) 0 (0.0) Moderate 0 (0.0) 0 (0.0) 0 (0.0) Severe 0 (0.0) 0 (0.0) 0 (0.0) Desaturations per hour Mean (SD) Median (min., max.) Seep time with SaO 2 < 90% (minutes) Mean (SD) Median (min., max.) max., maximum; min., minimum; SaO 2, arteria oxygen saturation. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 59

86

87 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Chapter 7 Acute oxygen substudy Background As there is some evidence that oxygen administration to patients with heart faiure might be harmfu, and that, in particuar, oxygen administration might be associated with adverse haemodynamic consequences, we undertook an acute haemodynamic substudy as part of the HOT tria. The evidence for harm with hyperoxia is imited to sma-scae studies, the main ones being by Haque et a. 21 and Mak et a. 22 Haque et a. 21 studied 10 patients with cass III and IV CHF who breathed 100% oxygen for 20 minutes. The administration of 100% oxygen reduced cardiac output from 3.7 /minute to 3.1 /minute and increased pumonary capiary wedge pressure (from 25 mmhg to 29 mmhg). Systemic vascuar resistance increased from 1628 dyn s/cm 5 to 2203 dyn s/cm 5, and there was no significant change in pumonary vascuar resistance. In a smaer substudy of ony seven patients, simiar changes were seen in patients breathing 24% oxygen. Mak et a. 22 studied 16 patients with sighty mider stabe CHF (NYHA cass II and III) and 12 subjects with norma LV function. Subjects again received 100% oxygen for 20 minutes. In the patients, LV end-diastoic pressure (equivaent to pumonary capiary wedge pressure in Haque s study 21 ) increased from 21 mmhg to 25 mmhg, cardiac output fe from 4.6 /minute to 4.1 /minute and SVR increased from 1626 dyn s/cm 5 to 1901 dyn s/cm 5. There were simiar changes to cardiac output and SVR in the contro subjects. Other evidence for the haemodynamics effects of oxygen is imited to very od studies of sma numbers of patients (such as a study of six patients 20 with acute myocardia infarction, in whom 100% oxygen caused a fa in cardiac output). There is no iterature suggesting harm from LTOT in patients with CHF. The oxygen dose that patients received in the main HOT tria was 2 /minute, equivaent to 28% oxygen and much ower than the 40% threshod for harm suggested by Haque et a. 21 There is no forma study assessing the effects of this dose of oxygen on haemodynamics in the type of patient incuded in the HOT study. We therefore assessed the haemodynamic effects of oxygen deivered for 10 minutes at a simiar concentration to that received by patients in the main HOT study (28%), with the aim of estabishing at east the short-term safety of ow-dose oxygen. Methods Incusion criteria To be incuded in the substudy, patients had to be undergoing cardiac catheterisation as part of the standard management of their CHF and meet the foowing criteria: suffer from heart faiure NYHA cass III/IV with LV systoic dysfunction confirmed by echocardiography (with < 40% or graded as at east moderatey impaired on visua inspection if an accurate ejection fraction coud not be cacuated) suffer from heart faiure from any aetioogy be receiving maximay toerated (and unchanged over previous 1 month) medica management of their heart faiure. be aged 18 years or over provide written informed consent and be abe to compete patient assessments. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 61

88 ACUTE OXYGEN SUBSTUDY Excusion criteria Patients did not undergo cardiac catheterisation merey to be incuded in the study. Patients were excuded if they: were unabe to provide informed consent had severe chronic airways disease FEV 1 /FVC < 70% and FEV 1 < 40% predicted and hypoxia (PaO 2 < 7.3 kpa or saturations < 90%). Procedures Cardiac catheterisation is a routine procedure taking around 20 minutes. The patient is usuay in hospita as a day case, arriving in the morning, having the test in the afternoon and going home in the evening. Patients were asked if they wished to take part in the substudy either at the time they were isted for cardiac catheterisation or on the day-case ward before the procedure, and provided written consent. As this was an observationa study, there were no randomisation or binding procedures. Before undergoing coronary angiography to assess coronary artery anatomy and eft ventricuography to assess LV function, patients underwent standard right heart catheterisation to measure: pressure in a cardiac chambers oxygen consumption oxygen content in pumonary artery and eft ventrice cardiac output pumonary and SVRs. Standard measures to ensure accuracy were as foows: the pressure recording equipment was zeroed at mid-chest, with rezeroing prior to each pressure measurement pressure readings were made at the end of norma expiration oxygen saturations were measured after caibrating the oximeter. After a compete set of data was acquired, the patient breathed 28% oxygen instead of room air (20.9% oxygen). After 10 minutes of oxygen breathing, the pressure, oxygen consumption and oxygen content measurements were repeated. Sampe size As there are few data avaiabe from previous studies, the sampe size was argey empirica. The study was stopped after 19 patients when it became apparent that oxygen had very few effects. Cacuations The intracardiac pressures and oxygen saturations were directy measured. Cardiac output was measured using the Fick principe from the oxygen consumption and the difference between the systemic arteria and pumonary arteria oxygen content, using the equation: O 2 content = Hb1:36SO 2 + 0:0032PO 2 : (1) The pumonary vascuar resistance was cacuated from (mean pumonary arteria pressure mean pumonary capiary wedge pressure)/cardiac output and given in Wood units. Simiary, SVR was cacuated from (mean systemic arteria pressure mean right atria pressure)/cardiac output and given in Wood units. 62 NIHR Journas Library

89 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Resuts Demographic data are shown in Tabe 35. The patients were younger than those in the main study, refecting the fact that most were undergoing assessment for possibe heart transpantation. The patients were we treated and had severe LV systoic dysfunction. Most had ischaemic heart disease and 32% were in atria fibriation. There were no compications during any of the procedures, but one patient experienced an episode of pumonary oedema 1 hour after catheterisation and was hospitaised overnight. TABLE 35 Baseine data for patients in the oxygen substudy Characteristic Baseine data (n= 19) Age (years) Mean (SD) 59.2 (14.7) Min., max. 27.1, 84.7 Mae (n) 7 Height (cm) Mean (SD) (5.9) Min., max. 159, 183 Weight (kg) Mean (SD) 89.5 (14.6) Min., max. 67,117 BSA (m 2 ) Mean (SD) 2.03 (0.18) Min., max. 1.69, 2.37 IHD/DCM (n) 15/4 NYHA III/IV (n) 10/9 ACEI/ARB (n) 19 BB (n) 16 MRA (n) 16 Loop diuretic (n) 17 (median 80 mg/day, range 0 200) Sinus rhythm/af 13/6 Haemogobin (g/d) Mean (SD) 13.8 (2.0) Min., max. 9.8, 17.3 LVEF (%) Mean (SD) 24.9 (10.2) Min., max. 5, 40 ACEI, angiotensin-converting enzyme inhibitor; AF, atria fibriation; ARB, angiotensin receptor bocker; BB, beta-bocker; BSA, body surface area; DCM, diated cardiomyopathy; IHD, ischaemic heart disease; max., maximum; min., minimum; MRA, mineraocorticoid receptor antagonist. The average dose given is the average daiy dose in furosemide equivaents, where 1 mg of bumetanide is assumed equivaent to 40 mg of furosemide. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 63

90 ACUTE OXYGEN SUBSTUDY Tabe 36 shows the effects of 10 minutes of 28% inspired oxygen on the directy measured intravascuar pressures during cardiac catheterisation. The patients had mid pumonary arteria hypertension and high eft heart fiing pressures but there was no effect of oxygen on any of the variabes measured. Tabe 37 shows the effects of 28% inspired oxygen on the oxygen saturation in pumonary artery and aorta together with the effects on derived haemodynamic variabes. The increase in the inspired oxygen ed to an increase in pumonary and systemic arteria saturation. There was a sma fa in pumonary vascuar resistance and an increase in cardiac output. The change in pumonary vascuar resistance correated with change in mean pumonary artery pressure rather than with pumonary capiary wedge pressure, and the change in cardiac output correated with change in pumonary artery saturation (Figure 10), suggesting that what sma changes to centra haemodynamics were seen were driven by changes in the pumonary rather than the systemic circuation. TABLE 36 Pressure measurements during right heart catheterisation Variabe Air Oxygen Paired t-test (p-vaue) Heart rate (beats per minute), mean (SD) (13.41) (14.04) 0.35 RAP (mmhg), mean (SD) 6.95 (3.36) 6.79 (3.95) 0.67 RV systoic (mmhg), mean (SD) (16.17) (17.82) 0.11 RV EDP (mmhg), mean (SD) 9.05 (3.60) 8.74 (4.62) 0.32 PA systoic (mmhg), mean (SD) (17.02) (16.86) 0.26 PA diastoic (mmhg), mean (SD) (7.46) (8.28) 0.53 PA mean (mmhg), mean (SD) (11.05) (11.04) 0.51 PCWP (mmhg), mean (SD) (8.88) (9.75) 0.17 LV systoic (mmhg), mean (SD) (19.56) (17.47) 0.12 LV EDP (mmhg), mean (SD) (7.46) (7.07) 0.78 Aorta systoic (mmhg), mean (SD) (17.29) (18.65) 0.85 Aorta diastoic (mmhg), mean (SD) (11.98) (12.33) 0.61 Aorta mean (mmhg), mean (SD) (12.78) (13.19) 0.70 EDP, end-diastoic pressure; PA, pumonary artery; PCWP, pumonary capiary wedge pressure; RAP, right atria pressure; RV, right ventrice. TABLE 37 Oxygen saturation and derived measures from cardiac catheterisation Variabe Air Oxygen Paired t-test (p-vaue) PA saturation (%), mean (SD) (9.77) (9.73) LV saturation (%), mean (SD) (1.89) (1.24) < Oxygen consumption (/minute), mean (SD) 0.24 (0.04) 0.24 (0.04) 0.56 PVR (Wood units), mean (SD) 2.78 (2.34) 2.12 (1.94) 0.02 SVR (Wood units), mean (SD) (4.78) (3.58) 0.07 CO (/minute), mean (SD) 3.89 (0.95) 4.16 (1.03) 0.03 Cardiac index (/minute/m 2 ), mean (SD) 1.92 (0.42) 2.06 (0.46) 0.02 SV (m), mean (SD) (16.80) (17.40) 0.11 CO, cardiac output; PA, pumonary artery; PVR, pumonary vascuar resistance; SV, stroke voume. 64 NIHR Journas Library

91 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 (a) 3 y = R 2 = Change in PVR (Wood units) Change in mean PA pressure (mmhg) (b) y = R 2 = 0.74 Change in CO (/minute) Change in mean PA pressure Change in mean PA saturation (%) FIGURE 10 Reations between haemodynamic variabes. CO, cardiac output; PA, pumonary artery; PVR, pumonary vascuar resistance. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 65

92

93 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Chapter 8 The home oxygen therapy tria quaitative substudy Background and rationae Despite a surviva benefit in COPD, patient adherence to LTOT is known to be probematic. 17,18 Adherence appears to be reated to symptom burden and performance status, and is better in patients who have had training in the use of the concentrator. 18,70 72 Given the ack of data about the use of home oxygen for peope with symptomatic heart faiure, this substudy expored the experience of patients and their informa carers regarding the use of HOT during participation in the main study to gain a deeper understanding of adherence issues than coud be gained from monitoring concentrator use aone. Breathess patients may have mixed feeings about oxygen therapy Positive feeings incude having treatment and an instinctive wecome of oxygen as it is routiney used during episodes of severe decompensated heart faiure by emergency heath professionas. The presence of an oxygen concentrator in the home and reguar attention from professionas may be reassuring. Patients with prior experience of cyinders may find the concentrator easier to hande with ess need for deiveries. However, the presence of the concentrator and associated paraphernaia can cause home disruption, imitations to activity, probems with tubing and anxieties about it going wrong. The fixed nature of the concentrator in the house may restrict outside activities and hence adherence is ower in patients abe to do activities outdoors. 74,75 In genera, adherence is ower for treatments given for maintenance or prevention than for those which are directed at an acute symptom with a cear temporay reated dose response reationship. In a RCT of oxygen of sham concentrators for peope with advanced disease and symptomatic breathessness, 50% of patients reported no benefit (for their breathessness) and did not want further oxygen therapy. 36,76 The sensation of breathessness correates poory with measures of ung function or arteria oxygen saturation and has a compex genesis in heart faiure. Two factors may make it difficut to identify the possibe roe of oxygen therapy for peope with advanced heart faiure, particuary if they are not hypoxaemic enough to warrant HOT in order to gain a surviva benefit. The first is the ack of immediate reief of breathessness (particuary when patients intuitivey beieve that they shoud be heped by oxygen) and the second is the fact that it might be a benefit that is mediated by airfow rather than by the oxygen. Aims and objectives This panned substudy aimed to expore the views of patients with CHF using a home oxygen concentrator, and their carers, with regard to benefits and burdens. In addition, we wished to understand their experience of this in the context of participation in the HOT tria. Methods Ethics approva was given by Northern and Yorkshire Research Ethics Committee (reference 09/H0903/43). Design This was a singe-centre nested quaitative study. A modified grounded theory approach using semistructured interviews was chosen, with the option of dyad or separate interviews with a carer, if there was one, rather than focus groups, to expore the experience of the participant within the context of their immediate famiy and home. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 67

94 THE HOME OXYGEN THERAPY TRIAL QUALITATIVE SUBSTUDY Samping strategy The participants were purposivey samped, in order to gain maximum variation, from patients in Hu recruited into the main HOT tria who had been aocated to an oxygen arm. Participants were samped for a varying time on the study: some within a few months, some after about 6 months and others after they had been aocated oxygen for a year. We anticipated that this approach woud aow for a spread of age, sex, types of iving arrangement (aone, with spouse, etc.), severity of CHF and comorbidities, which coud a potentiay affect a patient s experience of the oxygen concentrator. Potentia participants were identified and invited by the HOT research nurse according to the samping grid. Interested patients were given an information sheet and then contacted by the quaitative substudy researchers and an interview arranged. Written consent was obtained just prior to the interview. If participants informa carers wished to be interviewed as we, they received their own information sheet and signed their own consent. Data coection Background demographic data were coected from the baseine assessment of the main tria and incuded age, sex, severity of HF and ength of time on the HOT study. Semistructured dyad (if carer present) interviews then took pace with participants using a topic guide which was deveoped, based on the study aims, by the expert research group and informed by the pubished iterature about LTOT adherence and service users. Dyad interviews, in which patient and carer are interviewed together, have been used in a range of cinica speciaties, incuding paiative care, and can generate a richer understanding of needs and experiences than a singe interview. However, if the patient did not wish the carer to be interviewed, did not wish to be interviewed together with his or her carer or did not have a carer, he or she was interviewed aone. Participants were interviewed once in order to minimise burden. Interviews were conducted by Professor Miriam Johnson (MJ) and Dr Samantha Nabb (SN), neither of whom was invoved in seeing patients as part of the main study. The interviews were audio recorded and transcribed verbatim. An iterative approach was used; transcriptions were read, and any issues presented by patients and carers which were not incuded in the topic guide were added so that they may be specificay addressed in future interviews. Participants were interviewed at home or in the hospita cinic according to preference as ong as confidentiaity coud be maintained and the setting was as free as possibe from potentia interruption. Fied notes were taken by the researcher during and immediatey after the interview to record any observations about the body anguage of participants or the physica environment, particuary in reation to the presence of the oxygen concentrator. Data anaysis Thematic framework anaysis was used. This is an approach deveoped for conducting appied quaitative research, 77 and invoves moving through the stages of famiiarisation with the data, identification of a thematic framework, indexing the data using the thematic framework, arranging the data into charts, mapping and interpretation of the data. SN, MJ and Lesey Jones carried out the preiminary coding of interview transcripts, checking each other s coding for simiarity to increase rigor. SN used the NVivo computer software (QSR Internationa, Warrington, UK) to manage the data. A transcripts were coded by at east two researchers. The coding strategies and emerging themes were then agreed between a three researchers. Once the thematic framework had been drawn up, the findings were then synthesised to provide a summary of patient and carer views with particuar regard to the perceived benefits and burdens of HOT using a concentrator, and their experience of participation in the HOT tria. Data from informa carers and patients were anaysed together, as we considered the dynamics of the famiy unit to be integra to the use of oxygen therapy in the home. 68 NIHR Journas Library

95 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Resuts Participants Characteristics of the participants are summarised in Tabe 38. Two had stopped the oxygen therapy by the time of the interview but had continued providing data for the tria. A were interviewed in their own homes except for one who chose to be interviewed in the research cinic whie attending for an assessment for another study. Most who ived with a carer chose to be interviewed with them. Interviews took between 20 minutes and 1 hour. TABLE 38 Patients in the quaitative substudy ID Age (years) Sex KPS score Comorbidites Lives aone/with Interviewed with carer (if appicabe) Months in main study prior to interview 1 75 M 80 BiV PPM + ICD, DCM, hyperchoesteroaemia, IDDM, IHD With wife and daughter Wife and daughter M 60 AF, CABG, IHD, osteoarthritis, PVD, T2DM With wife Wife F 60 AF, arthritis, ICD, IHD, MI Aone n/a M 50 Anaemia, BiV PPM, CKD, HTN, T2DM 5 73 F 60 AF, CABG, ICD, MI, PVD, T1DM Aone n/a 3 Aone n/a 4; stopped O M 70 CKD, gout, HTN, MI With wife Wife F 60 Arthritis, HTN, IHD, T2DM With daughter Daughter M 60 AAA, AF, asthma, BiV PPM, IHD 9 85 F 60 COPD, IHD, trigemina neuragia M 60 Anaemia, BiV PPM + ICD, CABG, MI, VT M 50 AF, BiV PPM + ICD, COPD, IHD, T2DM With wife Wife 6 Aone n/a 9 Aone n/a 12; stopped O 2 at 6 With wife Wife F 60 CKD, IHD, MI, T2DM Aone n/a F 50 AF, BiV PPM, COPD, IHD, MVR With husband Husband and son F 60 BiV PPM, CVA, IHD With husband Husband M 60 AF, CABG, depression, IBS, ICD, IHD, HTN M 50 BiV PPM, Ca bowe, COPD, T2DM With wife Wife 7 With wife Wife F 60 AF, BiV PPM, CKD, HTN, IHD, Aone n/a 7 MI AAA, abdomina aortic aneurysm; AF, atria fibriation; BiV PPM, biventricuar permanent pacemaker; Ca, cancer; CABG, cardiac bypass graft; CKD, chronic kidney disease; CVA, cerebrovascuar accident; DCM, diated cardiomyopathy; F, femae; HTN, hypertension; IBS, irritabe bowe syndrome; ICD, impantabe cardioverter device; ID, identification; IDDM, insuin dependent diabetes meitus; IHD, ischaemic heart disease; M, mae; MI, myocardia infarction; MVR, mitra vave repacement; n/a, not appicabe; PVD, periphera vascuar disease; T1DM, type 1 diabetes meitus; T2DM, type 2 diabetes meitus; VT, ventricuar tachycardia. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 69

96 THE HOME OXYGEN THERAPY TRIAL QUALITATIVE SUBSTUDY Findings The agreed thematic framework used in the anaysis is presented as major and subthemes arising from the data (Tabe 39). The themes and subthemes are discussed in the foowing text with iustrative quotes ony for sake of brevity. Further quotation data for each theme can be provided if requested. Major theme 1: sense of sef and effect of chronic heart faiure The first area we examined was to understand the experience of iving with CHF, incuding participants understanding, their imitations and their ways of managing their condition in order to gain background for how an additiona factor, HOT, woud be received. Two reated major themes were apparent: a sense of sef and how this had been affected by the CHF, and the story (narrative) of participants ives in the context of their medica condition. Sef Participants taked about the need to sti be themseves, both as an individua and in terms of famiy reationships, despite the chaenges of iving with chronic i heath. Maintaining normaity as much as possibe was part of that aim. When maintaining their notion of sef became more difficut to do, because of increasing imitations, it was a source of distress. Intrinsic persona quaities and externa supports were ceary important in this process. TABLE 39 Major themes and subthemes Major theme Sef Subthemes Coping active adapt accept famiy humour Biographica disruption and sense of identity Ambivaence Narrative Iness interpretation comorbidity experience of heath-care professionas Life space incuding socia isoation physica imitations Heath beiefs threat/beiefs of others Impact on others Life narrative 70 NIHR Journas Library

97 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 TABLE 39 Major themes and subthemes (continued) Major theme Tria Subthemes Motivation persona support monitoring of condition access to technoogy infuence of famiy beief in specia treatment atruism Practicaities of tria of intervention noise tubes nose maintenance effect of socioeconomic status expectations Adherence with intervention Benefit burden baance short term vs. ong term O 2 as a reiever peace of mind/cam-me-down Incorporation into daiy routine roe of/effect on famiy Stigma both internaised and externa reating to sef-image Coping Participants coped using a variety of strategies, in particuar adaptation, humour (often inked with stoicism) and famiy support. I try to not et it affect my everyday ife. I just do what I want to do, within reason. You just have to cut your coth, and that s what I ve done. It is frustrating, it is frustrating, cause I ve aways been a person that just got on and got it done and, and I can t do that any more. But I don t dwe on what I can do, what I can t do, I actuay just think about what I can do, and I, it coud be a ot worse. Participant 17, woman Wife: And they [famiy] a come... Patient: They keep us going, that s my beief anyway... Wife: Yeah, they do. Participant 11, man with wife Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 71

98 THE HOME OXYGEN THERAPY TRIAL QUALITATIVE SUBSTUDY Often the coping strategies were ways in which participants had coped with adversity for many years and had served their purpose we, heping to preserve their identity as someone who coud manage. It s a basted nuisance [aughter]. It, but it doesn t, it rarey makes me fee i, I may be out of breath, I may be fatigued, but inside my head I m exacty the same person that I was before... But heart faiure, one ives with it, one ives round it. Participant 5, woman (withdrew from oxygen) However, sometimes, coping was overwhemed by i heath, particuary for those who saw themseves as the person who heped and organised others, rather than as the person receiving assistance. Famiy supports were tested and atered as a resut of the effect of the patient s condition, changing the reationship and roe within that unit, and not aways constructivey. Biographica disruption When coping strategies were overwhemed or unhepfu, and patients were unabe to adapt or accept the change in circumstances, the oss of confidence experienced with socia interaction (both in the home and more widey) was refected in a oss of sense of sef. The term biographica disruption has been used to describe this resut of chronic iness and disabiity. So heart faiure is a basted nuisance in the fact that it prevents John and I doing together the things that we, we, if we weren t, if I hadn t got it we d be going on raies, we d be going abroad, we d be visiting our friends in the south of Engand and our friends in Scotand. But it is difficut because you are sociay unacceptabe if you cannot, and our friends in Ireand, if you cannot spend a day awake, you can t, have to have 2 hours gone in the afternoon. Participant 5, woman I was the driver. And of course the work did take me round a ot but on, in driving... Yeah, and, and now, you know, I m totay dependent and it does, it does affect one, you know. Participant 15, man But I m very independent, I keep to mesef a ot, I am maybe me own worst enemy, but there you are... I was very active, I was aways the carer, I ooked after other peope... Participant 9, woman Major theme 2: the story (narrative) of the patient s ife in the context of his or her medica condition Narrative A participants reated their experience in a time-framed story of their ife, their iness and its effect in terms of the shrinking word, or space within which they ived their ife, which was imposed on them and their famiies. This story deveoped and contributed to beiefs about CHF and its treatment and about heath professionas, beiefs that were hed by themseves and their famiy members. Most had stories of very serious episodes of iness. Cause I did have a sight heart attack when I ived in [town], that were in 93, and then some, when I moved here, oh it was 2008, cause I were in [supermarket] at [town] and I had a massive heart attack and I coudn t move or anything, sweat poured off me, and a me imbs went numb, and when they took me to hospita I d had a cardiac arrest, whatever they ca it, and I ve never been right since, so... Participant 7, woman 72 NIHR Journas Library

99 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 We I ve had three heart attacks that they said they were heart attacks, and then two that were, they were mm er, mm um, mm, so five in a. I ve had a stroke, I m diabetic, insuin dependent, I ve had asthma since 60s, ate 60s, I ve had an underactive thyroid for 30/40 years, so. And I ve got a defib [defibriator] and a pacemaker, so. Sometimes it, you get a bit fed up about it, but I, I think of it this way, at east I m around. Participant 12, woman Part of the iness narrative incuded the infuence of both good and bad experiences of heath services in genera and heath-care professionas in particuar.... when I first had the heart attack I was, I was tod, quite, quite wrongy by a, a very young medic at one of the outpatients appointments. I d had an angiogram and the, the news was bad, that I d shortened my ifespan, there was too much damage to operate. So at the outpatients appointment I asked the, the doctor if he coud carify then, the phrase, what was it? It s shortened your ifespan. And he said Oh you be a right for 2 to 3 years... Participant 3, woman... so Professor [name] get me right again... me and Mary, never mind doctors, what, what, Professor [name] that, that side s, because they re, they re monitoring, they re doing it, you know, but when you get caed into your own doctor after you ve had a bood test saying your, your, your, your, your whatsits is a airy-fairy and a over the pace and they start knocking tabets off and they have, they, and they re doctors that don t know your history, you know, you know? Participant 4, man... I can t, I can t compiment the, the, what can I say, the care that I ve had from the hospitas and the [hospice] and the [academic heart care cinic], a, a of them have been fantastic for me... and whenever I ve gone in hospita its unbeievabe how, how, how the care has been for me and I m very gratefu for that. Participant 8, man Wife: Yeah. I found, and then he went on to the, the [accident and emergency] ward. That was just awfu, they were so, and I, and I did say it, it was so bad, I mean, in so much as they were aowing him to sef-medicate and they were aso medicating him, so his bood pressure, he didn t know whether he was on this earth or Fuer s earth. Patient: Fuer s earth [aughs]. Wife: It was a reay frightening experience, but then that s when you went from there to [inpatient hospita ward], wasn t it? And then it was, it was ike an entirey di, it was ike, we it was, it was just amazing, the difference was just amazing... Participant 16, man Sometimes I have breathing probems, but the a, the ady at the [hospice] gave me some exercises to do and she aso gave me one of these battery operated fans, which if I, if I get any probems with me breathing it does hep... Participant 8, man Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 73

100 THE HOME OXYGEN THERAPY TRIAL QUALITATIVE SUBSTUDY In genera, there was overt recognition of the imited treatment options at this stage of their iness. I know that I am getting towards the end now, there s no way I ve got much onger to go, so what do you do? Do you, you know, stop the heart tabets or do you con, continue with your diabetic fuctuations? And I m happy to do that. I ve had it for such a ong time. Participant 5, woman (withdrew from oxygen) But everybody, everybody I ve, I ve come across, ike even [Professor] he says You re on the right medi, there s not he says there s ony oxygen eft for you after, after that. You know, there s nothing... Participant 2, man Participants aso gave a narrative of their ife as part of maintaining sense of sef, trying to adapt and incorporate the imitations due to their iness with varying degrees of success.... but then things ike domestic chores, everyone tes you as they get oder that you reaise it s a,it s a tota waste of time because [interviewer aughs] tomorrow s as bad as it was today before you did it, so... since I retired, got a ot of increased hobbies and I, I, I centre on those reay, as being my priorities, because those are the things I enjoy doing. I can pick up and eave off as I want... Participant 3, woman I didn t run a marathon but just, just quicky, I was a exceent 800 metre runner... Mm. Lan, Lancashire champion... Trias for Engand. Yeah, I was very athetic. Participant 16, man I m a very independent person. My husband was 37 years in the Roya Air Force and I traveed the word with him, but I had four chidren and I took them with me. Many the times I ve traveed on my own with four chidren and he d meet me at the other end, you know, if I was ucky. So I m very independent but I ve ost me, me independence with me heart troubes. Participant 12, woman Wife: He s not a big one for waking. Patient: I m not a big one for waking now. Wife: But he, he was a big one for waking when he was Patient: years in the prison service. Wife: We yeah, he... Patient: 20 of those years were a dog-hander... Lot of waking invoved there. Participant 16, man The second area was that of the participants experiences of the tria and issues reating to adherence with oxygen as major themes. We ooked particuary to see if issues reated to the tria had an infuence over and above the issues reated to oxygen as possibe therapy but these were not apparent. Therefore, we report them as two separate themes. 74 NIHR Journas Library

101 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Experiences of the tria Tria Issues reating to the tria coud be divided into two subthemes: motivation for participating in the tria and factors heping or hindering continuation in the tria. Factors continuing with the tria reated soey to perceived burdens or benefits of the tria oxygen therapy and are discussed under adherence with oxygen, as we were unabe to identify whether or not the fact that patients were receiving the intervention as part of a tria (rather than for an expected cinica benefit) made a difference to their decision to continue with the oxygen. Motivation I d say give it a go, give something back, it, it s no imposition, and it is vountary, you can stop at any time. If you don t ike it when you put it on you can aways take it off, it s not, nobody s weding it to your face. Participant 2, man They aren t gonna cure my heart faiure now, it s too far gone, it s, you know, too far and it hep somebody ese get some comfort, I don t mind at a. Participant 12, woman Motives to enter the tria were mixed. Athough atruism was a feature, and expressed ceary by severa participants, there was a strong perception that participation in a tria gave the patient better support (e.g. from the research nurse and cinica academic team), better cinica management (e.g. more chance of being reviewed reguary by a senior doctor, if not the professor himsef), and access to the atest technoogy/treatments.... he said I d been painted a very back, back picture and woud I consider going to the academic cardioogy because that way I woud be monitored on a reguar basis and they, at the same time, woud use my, my records, etc., to their advantage, and I woud, it was mutuay beneficia. Participant 3, woman There was aso a significant infuence from the beiefs hed by the famiy carers, some amost instructing the patient to participate, for the reasons given above, especiay that this was seen as the best way of getting specia treatment, with one participant s wife seeing it as a way of getting her husband reinstated at the professoria unit, as his cinica care had been transferred to a district genera hospita. Wife:... because in the hospita in [district genera hospita] he was ike this, every day came a new doctor to say Ah, we re going to put a pacemaker on you. And next day came another doctor and said No, we re going to do this to you. And then we have three different... But the cardiac doctors, they are a over the pace, they don t know what they are taking about, and they make us very insecure. Participant 15, man I ve joined every study that I ve been offered, since, that I was eigibe for, and so when the oxygen came up my automatic reaction was if I m the right person then sure, you know, I m game for anything... And ike this tria, I mean I know they re going to send for me at reguar intervas, during the course, so. So no, I, I just get invoved in, in anything they ask me, with peasure reay. Participant 3, woman Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 75

102 THE HOME OXYGEN THERAPY TRIAL QUALITATIVE SUBSTUDY I ike these trias. The reason I ike em is they keep their eye on yer. Their eye on summat... cause sometimes if you re just in the routine system, you know, it can be 4 months, it can be 12 months before... Participant 6, man There were some specific comments reating to the initia entry on the tria, with some participants commenting that the oxygen concentrator was arger and noisier than they fet they had been ed to beieve and that the initia disruption in the house to fit tubing was more than they had expected. However, that in itsef did not have an impact on perseverance with the study at that stage, and the initia invitation discussions with the site investigator were recounted with good humour.... the machine is, in itsef, is quite noisy and, again, Professor [name] thinks it s just a mid hum, so I said We I, have you actuay istened to one? Participant 3, woman So I said, I said, Fancy getting you again [Interviewer aughs]. So he [Professor] said Oh, come on, what s the matter? So I forgot to say about the cyinder, so I said You didn t te me that it had to be piped downstairs as we so that you d access to the oxygen. He, I d made him augh because he said that there was gonna be 50 yards of tubing from here to kingdom come, from the machine upstairs, you see, which again coud ve been an accident waiting to happen. And he said Oh I must remember to ask peope, next time then, if they ive in a house or a bungaow. So I don t think it had entered his head that, you know, that the piping woud have to go through your was and down your boards, etc., but... Participant 3, woman And I was rather shocked [aughter] by the size of it, cause Professor [name] said it s the size of a cardboard box. Participant 17, woman Adherence The overwheming majority of the issues infuencing adherence were those practica issues aready we documented: noise and heat from the concentrator and a wide range of tube-reated chaenges. Tubes caused practica obstaces such as sore nose or ears, tying peope up, coming adrift (providing exceent oxygenation of the participant s ear or face) in bed at night or disrupting the stairway. Participants varied in their response to these difficuties within the context of how they managed their chronic condition overa and depending upon their character. Athough famiy and friends were seen by a as a source of support, participants who had maintained a sense of their roe within the famiy and of themseves more easiy incorporated the presence of the oxygen concentrator and tubing within their everyday ife with reative ease. Wife: The famiy come in and say Have you used that [oxygen] today? you know [aughs]. No, I haven t. We you shoud do... Participant 11, man However, those who seemed to be strugging with adapting to their current circumstances and sti spending a ot of energy in keeping up appearances, or who were finding it hard to accept how ife was, were acutey aware of the presence of the concentrator, activey keeping it out of sight, out of the way of 76 NIHR Journas Library

103 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 famiy and friends to the extent that if the grandchidren were visiting, they did not use it, ostensiby so as not to concern the grandchidren or to prevent adverse comments which they found embarrassing and stigmatising. The, we my missus, she s got ears ike a hawk [interviewer aughs] you know, and she woudn t put up with it, so... nobody taks about it, they know I m on it but they, they, they, they don t want to know... We my missus gets on about it saying How ong is that binking thing going to be there? but, you know. Participant 4, man Yeah, my, my granddaughters come, sometimes they come for a weekend, and Oh that machine, nan they say. So sometimes I switch it off and when they ve gone I (...) then, you see. But I m ucky, I suppose, I ive on me own so I can cope with it... We they re young aren t they? [interviewer aughs] It s an hindrance to them [aughs]. Participant 9, woman Such participants appeared to be ambivaent about the oxygen concentrator and tubing, stating first that it was no probem, but then immediatey presenting issues that they found difficut.... apart from it s another wire down the skirting board and everything ese, it s reay not a probem, except, to switch it on I have to go upstairs, and ikewise, downstairs to switch it off, back up again... it becomes a major ordea of having to ock the back door, go upstairs and ay on the bed, or switch it on there and then come downstairs. So that is the ony inconvenience for, at home, which isn t reay such a bad thing. Participant 3, woman One participant in particuar demonstrated an ambivaent attitude, trying to be positive, but then describing amost simutaneousy catastrophic responses to her condition and her current ife incuding the oxygen therapy. The oxygen concentrator had caused some married coupes to start seeping in separate rooms because of the disturbance of the noise. But, as I say, it makes a ot of row and, you know, I ve got to seep in a back bedroom... Participant 6, man So I shifted mesef into the back bedroom, eft [wife] in the front bedroom... No, no, don t want to upset [wife], so I stay in the back bedroom. Participant 16, man Some commented that this was a reasonabe option for them as they had a spare bedroom, but they fet that it woud be difficut for more deprived famiies for whom this might not be possibe. I don t have it in the room I seep in. I, I m ucky because I have three bedrooms... and I ive on me own, you see, so I put it in the other bedroom and cose the door. I worry about if the neighbours can hear it, but nobody s said anything now so they can t do. Participant 9, woman We put the machine in the dressing room... very fortunate we have a dressing room and the machine was in the dressing room and we coud sti, I coud, [husband s] a bit deaf but I coud sti hear it. Yes, it is noisy, yeah. Participant 5, woman (withdrew from oxygen) Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 77

104 THE HOME OXYGEN THERAPY TRIAL QUALITATIVE SUBSTUDY Others had found other ways of deaing with it (with one famiy pacing the concentrator in the haway during the day and in the ounge overnight with the door shut). This famiy came to that arrangement to avoid disturbing their adut daughter, as they were both hard of hearing and this was not a probem for them.... the ony thing is when I do, I run it through the door, I eave the machine there... so we can, if it... we can shut that door if we want to cut the noise down a bit and put the piping across here [gesticuating behind the sofa]. A at night-time, we fetch the machine, in, in here [ounge] and then we can cose up that door neary to. We re in the bedroom aong on the ground foor, so we go in there, put our door cose and...cause when we had it out there, [daughter] s upstairs and she coud hear it, so we cut that down and it heps her we can t hear it. Participant 1, man Conversey, one wife commented that the oxygen had heped her to seep at night because her husband s breathing was more setted at night than it had been when he was not taking oxygen. The concentrator was a significant imitation to mobiity which had some serious consequences. Going away on hoiday either became impossibe or had to be reframed. One famiy had sti managed to find a way to go away on hoiday by using a faciity owned by the oxygen company to rent out to cients but, athough they appreciated the opportunity to be away, they sti fet that it was restricting. A found the oxygen instaation and maintenance service to be exceent, responsive and reassuring. There were some concerns about the risk of fire, but these were in reation to the back-up oxygen cyinder which came with the concentrator and did not appear to be a major concern, mentioned ony by one adut daughter. Despite specific enquiry, concerns regarding smoking were not eicited. One participant stopped the oxygen therapy after an acute intercurrent iness which necessitated admission to a coronary care unit. She was severey i and unabe to speak for hersef. However, she remained aware enough to fee that, athough it was cear she was taking the oxygen as part of a cinica tria, the presence of HOT was perceived by the attending heath-care professionas (in particuar the ambuance staff and coronary care doctor) to be a marker for end-stage disease and that she woud be not for cardiopumonary resuscitation. When she was sufficienty recovered to go home, she was so disturbed by this, and worried that shoud a simiar event happen she woud be abeed as not for treatment, that she opted to stop the oxygen therapy athough she continued to provide data for the tria. In the hospita, they were handing me on to somebody ese, obviousy, and one of them said She s on home oxygen. And I said Just a minute. It s just a cinica tria. But I don t think that bit was heard... But one of the doctors in the intensive care unit, I don t know whether he thought I coudn t hear him, but I coud [aughs]. So the things he was saying at the end of the bed was that it s end-stage heart faiure and end-stage COPD... It affected their judgement of me, but I fet that I reay needed the oxygen that morning, because I was so breathess, and as soon as I put the oxygen on I was obviousy ess breathess because I d got the oxygen on. But I reay fet that the perception of me, as a patient, was, was competey wrong, as soon as they reaised I had, I, because as they waked in, I ve got the... [oxygen on]... Participant 17, woman (withdrew from oxygen) Athough some participants were very strict with attempting to fit in the fu 15 hours of therapy by panning hours of oxygen around the other things they wanted to do, most freey admitted to poor adherence. This was often inadvertent, simpy forgetting to use it if they were busy going out, or doing things around the house. Others deiberatey chose to use it ony at night, or when they did not have visitors or if they fet more breathess. Most admitted that they iked the idea of the concentrator in the 78 NIHR Journas Library

105 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 house as a safety net shoud they have a bad day with breathessness, feeing this might prevent them having to ca the emergency services for the oxygen that had been given on previous occasions. On the other hand, there was a concern that they woud become too reiant on it, which contributed to a feeing that they shoud restrict its use. But, you know, I m, I find it reassuring to have it, though sometimes I, I fee I might be dependent on it too... Participant 15, man Peace of mind I think, because I ve aways found when I have had the, the very, very bad uncontroabe angina, the first thing they do is give me oxygen, it cams you down... And it, it dus the pain knowing that it s there. If it, I ve got a rea bad go of angina I coud shove it on whie we re waiting for the ambuance peope to come... Participant 15, man... and then they ve brought me on the machine to, to give it a go, and it heped me a ot... I weren t gasping for breath, I weren t emergency, but it did cam me down, just cam me down, and after about haf an hour I d, I were back to norma... Yes, I don t want to get too reiant on it, so that if I do, don t go on it I haven t any probems; but it heps me a ot, it heps me seep and; this is what I ike about it reay, it cams me breathing down and it heps me seep through the night. Participant 14, woman There is a strong ayperson and heath-care professiona beief that oxygen is an important treatment for acute breathessness, which was a major barrier to adherence to a 15-hour-per-day regime. Many of the study participants had experience of repeated heart faiure decompensations and other acute events necessitating unschedued heath care for which administration of oxygen is a common first intervention. Thus, the use of oxygen for ong periods of time during the day and overnight, when at rest and when their breathing was usuay comfortabe, was difficut for many to understand. Wife: Switch it off, we have breakfast, and he goes a day without it, don t yer, ti... think, to be reay truthfu we didn t think we needed that, because he s not that poory that he had to go, oh me oxygen, he never puts it on even... Patient: No I ve never ever. Wife:... and he s busy during the day aren t, you know it... Participant 8, man This was further compicated by the beiefs of famiy members. One participant s wife had ambuatory oxygen for exertion-induced desaturation. This was immediatey beneficia for her and made intuitive sense. She and her husband found it hard to grasp why restricting mobiity with oxygen tubing for hours when one is comfortabe at rest, for exampe watching the teevision, shoud have any benefit especiay if the patient coud not perceive any immediate benefit. Those taking nocturna oxygen seemed to find adherence easier as it became part of the rhythm of going to bed. In addition, some (or their spouses) fet they sept better (when not having tube troube). We I think it heps me with me breathing at night-time, and I can have it on daytime if I think I need it, but I don t often need it. Participant 13, woman Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 79

106 THE HOME OXYGEN THERAPY TRIAL QUALITATIVE SUBSTUDY In genera, there was no cear pattern of adherence becoming easier or more difficut over time, athough some who had been on oxygen for some months commented that they fet that they were steadiy improving and put the improvement down to the oxygen. One participant described how he fet he had his roe back in the famiy again, as the famiy taxi driver, heping coect his grandson from coege. This heped probems seem ess significant in comparison with perceived benefit. Overa, patients weighed up the benefit burden baance of the oxygen therapy.... you get a bit, a bit sore and a bit resentfu of its irritation, to some extent. But then give it 5 minutes and you don t even know it s there reay. I think it s too eary days for me to say that it, it s beneficia. It, it appears to be going that way. But, as I say, it s, it s too eary days to say. I do get good spes but it s a ong time since I ve had a, a reay good spe so it coud be prompted by the oxygen. I woud ike to think it was, because I woud then opt to continue with it at some ater date. Participant 3, woman I ve not used a, a mask ong term so I don t know. I might have coped, after a whie got used to it and coped very we, you know. I mean I was a bit of wimp, I think, etting go so quicky. The physica, immediate, amost immediate physica probem took any ong-term thoughts from me, it was simpy that I coudn t manage with it, and if, ater, if you had said to me we know that oxygen therapy is good for you then I woud have said, fine, but I have to use a mask not these, yes. Participant 5, woman (withdrew from oxygen) As the concept of ong-term improvement in QoL (rather than specific and immediate symptom improvement) was difficut to accept, even for those who impicity trusted medica opinion, the practica and obvious burdens reating to equipment and restrictions were important. Again, the burdens were seen in the context of the participants personaity and abiity to adapt and accept their chronic iness. Thus, for some, the practica difficuties oomed arge, whereas, for others, the same issue was incorporated into everyday ife and was seen as a much smaer issue. Patient: See I m used to it now, it gets on [wife s] nerves a itte bit at times, but I can sit in here with it on... Wife: You don t hear it [aughs]. Patient:... with, with the tey, with watching the sport and so... Participant 8, man 80 NIHR Journas Library

107 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Chapter 9 Discussion The findings of the HOT study suggest that there is no symptomatic benefit at 6 months for patients with severe heart faiure from treatment with ong-term HOT when prescribed for 15 hours per day. Chronic heart faiure affects approximatey 2% of the adut popuation in the UK. Athough medica therapy has advanced dramaticay in the ast 20 years, patients with CHF tend to deteriorate with time and usuay eventuay die from the condition. As their condition worsens, so they become more and more breathess on minima exertion. In an attempt to reieve breathessness, HOT is commony prescribed, and yet there is no evidence that it is effective. The HOT study is the argest randomised tria to date of ong-term HOT for patients with CHF. The study is certainy needed; there is a proven mortaity benefit with HOT ony for patients with severe ung disease and hypoxia, yet most NHS prescriptions for home oxygen are given for indications other than ung disease. 39 HOT is not ony burdensome for patients but aso expensive to deiver. If it serves no purpose for patients in terms of symptoms reief, it shoud not be used. Cinica effectiveness Athough there was no difference in MLwHF questionnaire score between the LTOT and BMT arms at 6 months, we did see evidence for an improvement at 3 months. The LTOT group had a statisticay significant ower MLwHF questionnaire score (difference 5.47, 95% CI to 0.41; p = 0.03). This represents a cinicay important difference; 48 however, a difference of 10 was used to power the tria. The improvement may refect a pacebo effect and represent the effect of receiving active treatment, and having the oxygen avaiabe as a possibe acute treatment in times of severe symptoms. That the improvement disappeared at 6 months might be a resut of the waning of the pacebo effect couped with the burden of the oxygen usage. However, 3 months seems to be a ong time for a pacebo effect. Athough pacebo responses can ast up to 12 weeks, it is unusua and woud be particuary so in the situation where the patient is expecting an immediate benefit. There were twice as many deaths in the BMT arm (n = 12) as in the LTOT arm (n = 6), athough the impact of oxygen therapy on surviva was not statisticay significant (p = 0.16). Hazard rates were simiar up to around 6 months post randomisation, after which ony two more deaths were recorded in the LTOT arm but a further six were reported in the BMT arm. As we observed a significant difference in the MLwHF questionnaire score at 3 months (favouring LTOT), we hypothesised that LTOT might be keeping sicker patients aive onger and so be artificiay puing the MLWHF score up at 6 months (ower scores indicate a better QoL) reative to BMT. In the BMT arm, the higher mortaity of the more symptomatic patients woud have the effect that the average QoL of the group woud tend to improve as the more severey symptomatic patients died. To investigate this hypothesis, we repeated the primary anaysis mode restricted to patients who returned data at 3 and 6 months (and so necessariy were aive at 6 months). However, restricting the anaysis to those patients aive at a three time points showed no evidence of such an effect. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 81

108 DISCUSSION Quaity of ife measures Patients in the LTOT arm at 3 months reported an improvement in average breathessness over the past 24 hours on the numerica rating score by a cinicay important amount 57 (which was no onger apparent at 6 months). The improvement coincided with an improvement in the distress caused by breathessness, reporting of having coped better with breathessness, and a higher eve of satisfaction with the treatment received for breathessness. At 6 months, there was no difference in average or worse breathessness experienced over the ast 24 hours, but patients in the LTOT arm continued to have increased median scores for coping and satisfaction with treatment. There are two potentia expanations. First, it seems from other work that airfow (whether or not it is oxygen enriched) can reduce the sensation of breathessness; therefore, there may be a rea symptom improvement from using the fow of gas across the nasa mucosae, 36 which may not persist to 6 months, particuary in patients who are ess we. Second, the improvement in distress due to breathessness at 3 months may be because of the safety net issue seen in the quaitative substudy. In other words, patients may have had increasing sef-efficacy, that is the oxygen was something they had in reserve for a bad day. That effect, too, might disappear over time as the underying condition progresses or the therapy does not ead to hoped-for improvements. However, it is notabe that improvements in sef-rated coping with breathessness persisted for the first 12 months and satisfaction with treatment of breathessness unti 6 months. Safety There were no safety concerns with the use of HOT. There were theoretica concerns reated to possibe mechanica compications from tubing, and at a very eary stage there was a suggestion of an increase in admissions because of periphera vascuar disease in the oxygen-treated patients. However, as the study progressed, there were no significant differences in event rates between the two groups. Athough more participants in the BMT arm than in the LTOT and NOT arms experienced one or more SAEs, the difference was not statisticay significant. Might oxygen be associated with an increase in surviva? Athough the difference between the surviva curves is visuay striking (see Figure 8), the difference in surviva between the two groups was not statisticay significant. Couped with the resuts from the acute oxygen substudy, which did detect a favourabe haemodynamic response to oxygen, the evidence is reassuring that HOT, at east when the fraction of inspired oxygen is < 30%, is safe. However, the study was greaty underpowered to detect a difference in mortaity. The premature end to the study reduced the chances of detecting a surviva effect even further. It shoud be borne in mind that the target dose of oxygen at 15 hours per day is an amost arbitrary figure arrived at from anaysis of patients in ong-term trias of another condition. It remains possibe that oxygen used for shorter periods of time might be beneficia for peope with heart faiure. Why was the home oxygen therapy tria neutra? The most ikey expanation for the apparent ack of effect of LTOT for heart faiure in the HOT tria is that oxygen has no effect on breathessness in this cinica situation. There are severa possibe confounding issues. 82 NIHR Journas Library

109 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Patient seection Were the patients in the HOT study representative of patients with severe heart faiure? If the patients were insufficienty i at baseine, it is unikey that they woud benefit from treatment directed at paiative care. The study protoco was designed to ensure that ony severey symptomatic patients were recruited, based on their NYHA status. However, the NYHA scae is reativey crude; those in cass IV are breathess at rest (and are thus argey either bed-bound or in hospita) and therefore most of the patients recruited into HOT were cass III patients (that is they were breathess on mid exertion). It may be that this eve of symptom was simpy insufficienty severe to aow any benefit from home oxygen. However, the KPS gives finer gradations of performance status and show that amost a the patients needed hep with activities of daiy iving, consistent with a group of patients who were significanty imited by their condition. On average, patients recruited to the HOT study were neary 10 years oder than patients in most studies of CHF (whose average age is typicay around 60 years), suggesting that the patients were representative of patients seen in day-to-day practice. That the patients had severe heart faiure is confirmed by a number of the baseine variabes. Two-thirds had severe LV systoic dysfunction on echocardiogram, and the mean NT-proBNP at 4500 ng/ was very high. The 6MWT distance was grossy reduced, at 120 m. The average distance covered by a simiary aged popuation in 6 minutes is around 660 m, 78 but fas with age. 79 In an unseected popuation of patients with CHF of simiar age, the median (interquartie range) 6MWT distance was 345 m ( m). 80 Patients with a 6MWT distance of 120 m had an 11-fod higher risk of death than patients with a 6MWT distance above 376 m. 80 It does thus seem that patients recruited to HOT did have severe heart faiure. Prevaence of hypoxia The patients with chronic airways disease who gain a surviva advantage from home oxygen are severey hypoxic (PaO 2 < 55 mmhg 14 ). None of the patients in HOT was so severey hypoxaemic. There is a persistent beief that breathessness equates with hypoxia. Patients with treated chronic stabe heart faiure are commony assumed to be hypoxic at rest or during exertion, but the overwheming body of evidence demonstrates that they are not. 25,26,81 85 In the patients recruited to the HOT tria, arteria oxygen saturation was norma and there was no significant change in arteria oxygen saturation during exercise or during recovery from exercise. The HOT tria was designed to assess the prevaence of hypoxia as part of the SDB syndrome. Epidemioogica evidence suggests that SDB is very common in patients with heart faiure when a cut-off of 15 episodes of apnoea/hypopnoea per hour of seep is used as the diagnostic standard. 27 However, even though the patients in the tria had severe heart faiure, the mean AHI at baseine was ony 9.8 per hour, with more than 75% having an AHI beow the diagnostic threshod of 15. The ack of symptomatic benefit seen in the HOT tria may thus simpy refect the very ow prevaence of hypoxia at rest, during exercise or overnight. There may be a very sma subset of patients with severe hypoxia as a resut of CHF aone, but we did not detect such patients despite specificay targeting the most severey symptomatic patients. Adherence We found that adherence to the 15 hours-per-day prescription of home oxygen was poor. In the LTOT arm, the mean usage was ony 5.4 hours per day, with two-thirds of patients using the oxygen for < 8 hours per day. Ony a sma minority of patients used the oxygen as prescribed, athough average use between readings often fuctuated, indicating that patients often had periods when they used the machine more or ess. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 83

110 DISCUSSION These findings are broady in ine with reports from the use of HOT in patients with chronic airways disease (Tabe 40). There is a variety of reasons for poor adherence: There was equipoise in prescribing the oxygen: because there is no evidence that oxygen proongs surviva, the investigators coud not impress any mortaity benefit on the patients, perhaps essening the patients enthusiasm for oxygen compared with patients with chronic airways disease oxygen therapy. As the findings from the quaitative substudy suggest, oxygen therapy is certainy burdensome. The physica imitations imposed by the oxygen tubing and the symptoms caused by the oxygen deivery (dry nose and mouth in particuar) caused probems for some patients. Athough patients were tod that it was unnecessary to take oxygen with them (in cyinders) when they eft the house, some were concerned at stopping oxygen when they eft home. It s possibe that making oxygen cyinders avaiabe might improve adherence. Patients had mixed reasons for taking part in the study, some thinking that they woud receive better care when taking part in the study and who were perhaps ess we motivated to use the oxygen. It is certainy the case that patients taking part in studies have a better prognosis generay. 93 Oxygen did not necessariy cause an immediate improvement in patients symptoms: as time on treatment passed, patients became ess enthusiastic about oxygen use because of ack of perceived benefit, particuary in comparison with hoped-for benefit. TABLE 40 Studies of adherence in patients receiving LTOT for chronic airways disease Source Subjects Adherence Evans et a concentrator patients prescribed 15 hours per day Mean 13.3 (SD 2) hours per day Vergeret et a hypoxic COPD patients Fixed: mean 14 (SD 3) hours per day Washaw et a. 72 Randomy assigned to fixed unit ony or fixed + portabe 61 patients reassessed for use and prescription appropriateness Portabe: mean 17 (SD 3.5) hours per day 45.9% inadequate prescription 29.5% adherence with correct prescription Howard et a concentrators post use If prescription < 15 hours per day then 9.9 hours per day Compared prescription and concentrator cocks If prescription > 15 hours per day then 13.4 hours per day Restrick et a patients interviewed and foowed up 74% used > 12 hours per day Morrison et a LTOT patients (79% COPD): 3-year data Mean 14.9 (SD 6) hours per day 44% < 15 hours per day Granados et a LTOT patients (70% COPD) 31% met a criteria for adherence to adequate prescription 61% compiant Pépin et a COPD patients 45% achieved > 15 hours per day Compared prescription and actua use Prescription mean 16 (SD 3) hours per day Actua use mean 14.5 (SD 5) hours per day Ringbaek et a of 182 LTOT patients 65% had acceptabe adherence Ambuatory use positivey affected adherence Atis et a of 1100 patients responded to questionnaire 28.2% sef-reported use was > 15 hours per day Mean 9 (SD 6.8) hours per day 84 NIHR Journas Library

111 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 The poor adherence with the 15 hours per day prescription was the utimate reason that the HOT tria came to a premature end. It is possibe that oxygen may have conferred some benefit when used as short-burst therapy to reieve acute breathessness: however, there was no evidence of sustained improvement in symptoms with oxygen therapy. Athough there was some evidence of short-term reduction in breathessness intensity as measured by the NRS, and onger-term improvement in subjective coping with breathessness, we cannot be sure that these findings are attributabe to oxygen and not simpy to increased airfow over the nasa mucosae. Substudies Acute oxygen substudy (see Chapter 7) The major aim of the oxygen substudy was to demonstrate the safety of oxygen given at the eve to be used in the main HOT study. There had previousy been concern that high concentration of inspired oxygen might be deeterious in peope with heart disease, and there had never been any studies to show whether or not 28% oxygen had any haemodynamic effect. Some previous studies have suggested that there may be a sma fa in pumonary vascuar resistance with oxygen given at > 90%, 94 but other studies have suggested that 100% has no haemodynamic effect. 95 This is the first study to examine the effects of 28% oxygen on haemodynamic variabes in patients with heart faiure. We found no evidence of a deeterious haemodynamic effect of 28% oxygen. There was, in fact, a sma fa in pumonary vascuar resistance and a sma increase in cardiac output with oxygen, findings suggesting sma improvements in haemodynamics. The mechanisms behind the apparent improvement cannot be assessed from this experiment. There are two major imitations to the findings. First, the patient group was a highy seected one, namey patients having cardiac catheterisation for cinica reasons, in arge part for transpant assessment. Patients in the substudy tended to be younger than the patients recruited into the main HOT study, and tended to have ony a singe (cardiac) pathoogy. Second, we were abe to give the oxygen for ony a reativey brief period, and it is possibe that onger exposure to increased inspired oxygen may have had ater harmfu effects. The order of the two tests, namey air then oxygen, was not randomised and the patients were not binded to the administration of oxygen. Concusion The resuts of the acute haemodynamic substudy suggested that it is reasonabe to use 28% oxygen in patients with severe heart faiure and that there was no evidence that patients in the main HOT study were being exposed to a dangerous intervention. Quaitative substudy (see Chapter 8) Work in other areas of treatment adherence has resuted in the deveopment of the Common Sense Mode, which expains a patient s adherence to treatment as the combined resut of two main aspects of the way in which they perceive their iness: cognitive iness perception and emotiona iness perception. Information and understanding about their medica condition, the way in which treatment may hep and the aims of therapy are important in deveoping patients cognitive iness perception. For many, information comes from sources other than their heath-care professiona. Likewise, patients emotiona response to their heath probems wi resut from interpay between their ife experience, their psychoogica make-up and coping mechanisms. If heath-care professionas are not aware of such heath and persona beiefs, then attempts to improve adherence may be unsuccessfu. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 85

112 DISCUSSION The difficuties of LTOT, and patient concerns about overdependence, or even addiction, which contribute to poor adherence in cinica practice are described in Adherence and can be reated easiy to the Common Sense Mode. Athough the mode, and others simiar to it, refers to sef-reguation in the context of chronic iness and its treatment rather than in the context of a cinica tria for peope with chronic iness, we found no tria-specific issues which affected a participant s adherence to oxygen therapy. Even though there is atruism associated with giving tria consent, there was a strong beief that participation in a cinica tria was a way of getting access to the best management for their heart faiure. Once in the tria, patients adherence to oxygen therapy was heaviy infuenced by their beiefs about the therapeutic target of oxygen (as immediate symptom benefit for breathessness or for end-stage disease). Despite carefu expanation to the contrary in the tria patient information eafet, reinforced by discussion with the site principa investigators and research nurses, these beiefs were firmy hed by some, earned through past experience of their chronic iness, past heath-care interventions for that iness and ay wisdom. A further compication in some of those who did fee they were benefiting was a worry that they shoud become dependent upon the oxygen if they used it too much, even though there was expicit instruction to use it for at east 15 hours per day. The strong beief as to the purpose of oxygen in its expected reationship to immediate symptom reief ed to an apparent cognitive and emotiona dissonance, which resuted in some participants preferring to have the concentrator in the house but using it ony as a safety net for bad symptom days. The participants response to the practica chaenges and difficuties encountered by the oxygen mirrored their emotiona response to their chronic iness, how they coped with it and how they had functioned, or not, within their famiy reationships and reationships with heath professionas over many years. Thus, the same practica issues (noise, tubes), and same apparent sources of coping (e.g. famiy), resuted in different rates of adherence, which were consistent with the individua s pattern of approach to stressfu situations. The best adherence was seen in those, particuary those who were aocated NOT ony, who found a way to incorporate the oxygen therapy into the routine of their daiy ives (e.g. part of the routine of going to bed), and managed the disruptions to daiy ife in the home in a practica manner (e.g. moving bedrooms). Those who were unabe to be fexibe in their view of what was norma ife appeared to be ess compiant. Concusions Participants viewed study participation in the tria both as an atruistic act and as a way of accessing optima cinica care. Adherence or not in the oxygen arm did not appear to be specificay reated to the context of a cinica tria. There was a deep-seated beief that oxygen is a therapy for acute deterioration or for those with end-stage disease. Thus, the use of LTOT was counterintuitive despite cear expanation of the tria s aim.this misunderstanding formed a poor basis for subsequent weighing up by the participant of the benefit burden baance of the LTOT. In addition, the individua s psychoogica make-up, ways of coping, and previous iness and ife narrative infuenced their emotiona response to the burden imposed by oxygen therapy and affected whether or not they coud incorporate oxygen therapy into the rhythm of daiy ife. Those who coud appeared to use the oxygen more often than those who coud not. Limitations We were unabe to bind oxygen deivery as the origina study design had panned. However, the open tria was a pragmatic one which addressed the important question of treatment effectiveness. The tria was stopped eary because of poor adherence to the 15 hours per day prescription. The prescription was based on extrapoation from studies of patients with a different pathoogy, chronic airways disease, who had severe hypoxia. It may be that shorter periods of exposure might have been effective, in terms of either symptom reief or preventing hospitaisation. 86 NIHR Journas Library

113 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 The a priori sampe size cacuation was based on the independent-sampe t-test at 6 months, giving a conservative estimate of 200 participants to detect a difference in a MLwHF questionnaire score of 10 points, assuming a SD of 25 points (effect size of 0.4), at 80% power with 5% significance. Aowing for 10% attrition at 6 months, we needed to recruit and randomise 222 participants. The primary resut was the treatment effect at 6 months extracted from an adjusted covariance pattern mode. Adjusting for baseine covariates by using an anaysis of covariance (ANCOVA) increases the statistica power compared with a t-test for the same number of participants. With an ANCOVA, the sampe size required to detect the same difference can be reduced by a factor of (1 ρ 2 ), where ρ is the correation between the baseine covariates and the outcome. For exampe, with a sampe size of 200 at 6 months, assuming a moderate correation of 0.5 between the baseine covariates and MLwHF questionnaire score at 6 months, we woud have had 90% power to detect an effect of magnitude 0.4 with an ANCOVA compared with 80% power with the t-test. An ANCOVA can use data avaiabe ony for the time point of interest, here 6 months; however, in the covariance pattern mode, we coud incude the 3- and 12-month post-randomisation assessments. In a repeated-measures mode, missing data are ess probematic than when using a t-test, as observations at each time point infuence estimates of treatment effects at every other time point, owing to the specification of a covariance pattern for the within-patient repeated measures. Thus, patients whose observations are imited to the 3- and/or 12-month time points wi nevertheess be taken into account when the estimate at 6 months is made. Ceary, such individuas wi not infuence the estimates as greaty as individuas whose data are compete, but the phiosophy is that some data from a patient is better than none. The extra information further increases the power avaiabe from the ANCOVA; however, it is beyond our knowedge to quantify this extra benefit and we have found no iterature that discusses this exact probem. Cacuating a sampe size based on an adjusted repeated-measures anaysis is compicated and requires knowedge of parameters that are not known in advance (particuary the correation between baseine covariates and the outcome). The cacuations are sensitive to initia assumptions and so we used a t-test to provide a conservative estimate to minimise the risk of underpowering the tria. A tota of 88 participants provided usabe primary outcome data at 6 months; 83 aso had compete baseine data for MLwHF questionnaire score, age, NT-proBNP eve and creatinine eve. To detect a 10-point difference with a SD of 25 at 6 months, the estimated power avaiabe in a two-sided test of two independent means is 46% but by ANCOVA (adjusting for the covariates and assuming a correation between baseine covariates and 6 months MLwHF questionnaire score of 0.5) is 55%. As a resut of the tria being underpowered, the risk of a type II error is increased and any of the resuts, especiay those with an associated p-vaue greater than 0.05, are inconcusive. The ow sampe size means it is not possibe to distinguish whether the ack of a detected effect is a resut of the tria being underpowered or whether there is a true ack of effect. In addition, a ack of adherence with oxygen may have diuted any effect of ong-term home oxygen. The MLwHF questionnaire is the most widey used instrument for the assessment of QoL in patients with heart faiure. It asks patients to rate the severity of their symptoms on a 6-point scae (from 0 to 5) but does not assess the impact of those symptoms on the patient s ife directy. It might be that another scae, such as the Kansas City Cardiomyopathy Questionnaire, 99 woud have been more sensitive to changes in heath state. The Kansas City Cardiomyopathy Questionnaire asks patients to record the impact that a symptom has on their ife rather than merey the presence or absence of a symptom. It might be that the patients in the HOT study were sufficienty habituated to their symptoms that athough they were, perhaps severey, breathess, they were used to it and thus fet that the breathessness had itte impact on their ife. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 87

114 DISCUSSION Concusions The prevaence of hypoxia in patients with severe heart faiure, at rest, foowing exercise and during an overnight seep test, is ow. There is no evidence that LTOT, athough it is safe, improves the symptoms, prognosis or severity of heart faiure in patients with severe CHF at 6 months. There is no evidence to support the use of HOT in patients with heart faiure but, as the study was terminated eary, we cannot excude the possibiity of a type II error. Recommendations for future research We suggest that two further studies might be appropriate: 1. a tria of patients with severe heart faiure randomised to have emergency oxygen suppy in the house, suppied by cyinders rather than oxygen concentrator, powered to detect a reduction in admissions to hospita 2. a study of bed-bound patients with heart faiure who are in the ast few weeks of ife, powered to detect changes in symptom severity. However, given the probems with the conduct of HOT encountered in the present study, mounting such a tria woud face simiar probems with both binding (and the use of a sham device) and adherence. The HOT investigators are in cose contact with the investigators of the OXYGEN-HF tria (ACTRN ). OXYGEN-HF is a randomised tria of 285 patients comparing the effects of HOT, medica air (pacebo) and no treatment (contro) on a-cause admissions to hospita at 6 months. Many of the tria s secondary end points are the same as ours, and the findings of both studies wi be of vaue in a meta-anaysis. 88 NIHR Journas Library

115 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Acknowedgements We woud particuary ike to acknowedge the contribution made to the study by the patients. The tria was demanding for the patients in terms of both disruption to their ives (and to their houses) and data coection. We are particuary gratefu to Mr Patrick Fouk whose hep and advice was vita. Mr Fouk aso heped with drafting the Pain Engish summary. Thank you. The tria coud not have happened without the support of the recruiting centres, and we gratefuy acknowedge the support we received from each one. We particuary acknowedge Mr Dominic Feows and Mr Pau Atkin, and the nurses at the Hu centre, who provided a great dea of support to the other recruiting centres. The tria owes a arge debt of gratitude to Dr Ian Harvey, the tria manager, without whose energy and support the tria woud have been impossibe. Mr James Iingworth on behaf of the tria sponsors, the Hu and East Yorkshire NHS Hospitas Trust, aso provided invauabe support. We aso acknowedge the hep, support and advice of the HTA programme, particuary Mrs Lesey Dodd, the programme manager, who was unfaiingy hepfu and encouraging. The investigators have aways fet we supported by the HTA programme. The home oxygen therapy tria Chief investigator Professor Andrew Cark, Honorary Consutant Cardioogist, Department of Academic Cardioogy, University of Hu, Caste Hi Hospita, Cottingham, UK. Tria manager Dr Ian Harvey, Academic Cardioogy, Caste Hi Hospita, Cottingham, UK. Tria co-ordinators Sarah Cockayne, York Trias Unit, University of York, York, UK. Sara Rodgers, York Trias Unit, University of York, York, UK. Statistician Caroine Fairhurst, Statistician, York Trias Unit, University of York, York, UK. Co-investigators Dr Victoria Agar, Senior Lecturer in Medica Statistics, Hu York Medica Schoo, University of York, Hesington, York, UK. Dr Michae Greenstone, Consutant Chest Physician and Honorary Cinica Senior Lecturer, Medica Chest Unit, Caste Hi Hospita, Cottingham, East Yorkshire, UK. Dr Susan Griffin, Senior Research Feow, Centre for Heath Economics, University of York, York, UK. Professor Miriam Johnson, Professor of Paiative Medicine, Hu York Medica Schoo, University of Hu, Hu, UK. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 89

116 ACKNOWLEDGEMENTS Professor Jerry Murphy, Consutant Cardioogist, Darington Memoria Hospita, Darington, UK. Professor Iain Squire, Cardiovascuar Medicine, Department of Cardiovascuar Sciences, and Nationa Institute for Heath Research Cardiovascuar Biomedica Research Unit, Genfied Hospita, Leicester, UK. Professor David Torgerson, Director of York Trias Unit, University of York, York, UK. The study committees Tria Steering Committee Independent chairperson Professor Henry J Dargie, Honorary Senior Research Feow, University of Gasgow, Gasgow, UK. Independent members Professor Andrew Coats, Facuty of Medicine, University of Sydney, Austraia. Professor Martin Cowie, Nationa Heart and Lung Institute, Dovehouse Street, London, UK. Patient representative Patrick Fouk. Researchers Professor Andrew Cark, Professor Miriam Johnson and Dr Ian Harvey. Data Monitoring and Ethics Committee Professor Theresa McDonagh, Consutant Cardioogist, King s Coege Hospita, London, UK. Dr Roy S Gardner, Consutant Cardioogist, Scottish Advanced Heart Faiure Service, Goden Jubiee Nationa Hospita, Cydebank, UK. Dr Suzanna Hardman, Consutant Cardioogist with an interest in Community Cardioogy, The Whittington and University Coege London Hospitas, London, UK. Giian Worthy, Statistician, Keijnen Systematic Reviews Ltd, Escrick Business Park, Escrick, York, UK. 90 NIHR Journas Library

117 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Contributions of authors Professor Andrew L Cark (Professor of Cinica Cardioogy) conceived the study and appied for, and received, the origina grant. He was principa investigator for the study. He conducted the review of the data for HOT in heart faiure and is the principe author of this report. Miriam Johnson (Professor of Paiative Care) cowrote the initia grant appication, tria protoco (main study and quaitative substudy) and ethics appication (main study and quaitative substudy), was a member of the tria management group, provided paiative care and symptom outcome assessment expertise, was principa investigator for the quaitative substudy, conducted data coection and anaysis for the quaitative substudy and contributed to writing the report. Caroine Fairhurst (Statistician) conducted the statistica anaysis of the main tria, prepared the resuts for pubication, and contributed to the preparation and review of the fina report. David Torgerson (Director of York Trias Unit, Heath Sciences) heped design the study, gave tria methodoogica advice and was a co-appicant. Sarah Cockayne (Research Feow, Heath Sciences) was a tria co-ordinator on the study, and contributed to the design of the study and to the preparation and review of the fina report. Sara Rodgers (Research Feow, Heath Sciences) was a tria co-ordinator on the study, and contributed to the preparation and review of the fina report. Susan Griffin (Senior Research Feow, Heath Economics) anaysed the EQ-5D data, and contributed to the preparation and review of the fina report. Victoria Agar (Statistician) provided statistica support during the design and conduct of the main tria. Lesey Jones (Senior Lecturer, Socia Sciences) cowrote the quaitative substudy protoco and provided senior quaitative expertise regarding data anaysis and report writing. Samantha Nabb (Heath Psychoogist, University of Hu) heped to create the interview schedue investigating the patient and carer experience of the oxygen concentrator, interviewed patients and carers and anaysed the data. Ian Harvey (Tria Manager) contributed to the design, deveopment and deivery of the tria, and to the preparation and review of the fina report. Iain Squire (Professor of Cardiovascuar Medicine) heped design the study, contributed to the writing of the report and was a co-appicant. Jerry Murphy (Professor of Cardiovascuar Medicine) heped design the study, contributed to the writing of the report and was a co-appicant. Michae Greenstone (Consutant Chest Physician and Honorary Cinica Senior Lecturer) co-wrote the initia grant appication and tria protoco, and contributed to the writing of the report. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 91

118 ACKNOWLEDGEMENTS Pubications Cark AL, Johnson MJ, Squire I. Does home oxygen benefit peope with chronic heart faiure? BMJ 2011;342:d234. Davidson PM, Johnson J. Update on the roe of paiative oxygen. Curr Opin Support Paiat Care 2011;5: Gadoud A, Johnson MJ. What paiative care cinicians need to know about heart faiure. Progress Paiat Care 2014;22: Data sharing statement A avaiabe data can be obtained from the corresponding author. 92 NIHR Journas Library

119 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 References 1. Cowie MR, Wood DA, Coats AJ, Thompson SG, Pooe-Wison PA, Suresh V, et a. Incidence and aetioogy of heart faiure; a popuation-based study. Eur Heart J 1999;20: /euhj Heidenreich PA, Abert NM, Aen LA, Buemke DA, Buter J, Fonarow GC, et a. Forecasting the impact of heart faiure in the United States: a poicy statement from the American Heart Association. Circ Heart Fai 2013;6: Nationa Institute for Cardiovascuar Outcomes Research. Nationa Heart Faiure Audit Apri 2012 March London: Nationa Centre for Cardiovascuar Prevention and Outcomes, University Coege London; Ceand JG, Cark AL. Deivering the cumuative benefits of tripe therapy to improve outcomes in heart faiure: too many cooks wi spoi the broth. J Am Co Cardio 2003;42: Stewart AL, Greenfied S, Hays RD, Wes K, Rogers WH, Berry SD, et a. Functiona status and we-being of patients with chronic conditions. Resuts from the Medica Outcomes Study. JAMA 1989;262: Currow DC, McDonad C, Oaten S, Kenny B, Acroft P, Frith P, et a. Once daiy opioids for chronic dyspnoea: a dose increment and pharmacovigiance study. JPainSymptomManage2011;42: Abernethy AP, Currow DC, Frith P, Fazekas BS, McHugh A, Bui C. Randomised, doube bind, pacebo controed crossover tria of sustained reease morphine for the management of refractory dyspnoea. BMJ 2003;327: Jennings AL, Davies AN, Higgins JP, Gibbs JS, Broadey KE. A systematic review of the use of opioids in the management of dyspnoea. Thorax 2002;57: thorax Johnson MJ, McDonagh T, Harkness A, MacKay S, Dargie H. Morphine for breathessness in chronic heart faiure. Eur J Heart Fai 2002;4: Oxberry SG, Torgerson DJ, Band JM, Cark AL, Ceand JG, Johnson MJ. Short-term opioids for breathessness in stabe chronic heart faiure: a randomized controed tria. Eur J Heart Fai 2011;13: Oxberry SG, Band JM, Cark AL, Ceand JG, Johnson MJ. Repeat dose opioids may be effective for breathessness in chronic heart faiure if given for ong enough. J Paiat Med 2013;16: Medica Research Counci Working Party. Long term domiciiary oxygen therapy in chronic hypoxic cor pumonae compicating chronic bronchitis and emphysema. Report of the Medica Research Counci Working Party. Lancet 1981;1: Nocturna Oxygen Therapy Tria Group. Continuous or nocturna oxygen therapy in hypoxemic chronic obstructive ung disease: a cinica tria. Ann Intern Med 1980;93: / Cranston JM, Crockett AJ, Moss JR, Apers JH. Domiciiary oxygen for chronic obstructive pumonary disease. Cochrane Database Syst Rev 2005;4:CD cd pub2 Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 93

120 REFERENCES 15. Uronis H, McCrory DC, Samsa G, Currow D, Abernethy A. Symptomatic oxygen for non-hypoxaemic chronic obstructive pumonary disease. Cochrane Database Syst Rev 2011;6:CD Pépin JL, Barjhoux CE, Deschaux C, Brambia C. Long term oxygen therapy at home. Compiance with medica prescription and effective use of therapy. Chest 1996;109: /chest Avdeev SN, Aisanov ZR, Chuchain AG. Compiance as a critica issue in ong term oxygen therapy. Mondai Arch Chest Dis 1999;54: Katsenos S, Froudarakis ME, Charisis A, Vassiiou MP, Constantopouos SH. Long term oxygen therapy in Ioannina. Respiration 2004;71: Croxton TL, Baiey WC. Long term oxygen treatment in chronic obstructive pumonary disease: recommendations for future research: an NHLBI workshop report. Am J Respir Crit Care Med 2006;174: Thomas M, Mamcrona R, Shiingford J. Haemodynamic effects of oxygen in patients with acute myocardia infarction. Br Heart J 1965;27: Haque WA, Boehmer J, Cemson BS, Leuenberger UA, Siber DH, Sinoway LI. Hemodynamic effects of suppementa oxygen administration in congestive heart faiure. J Am Co Cardio 1996;27: Mak S, Azevedo ER, Liu PP, Newton GE. Effect of hyperoxia on eft ventricuar function and fiing pressures in patients with and without congestive heart faiure. Chest 2001;120: Moore DP, Weston AR, Hughes JM, Oakey CM, Ceand JG. Effects of increased inspired oxygen concentrations on exercise performance in chronic heart faiure. Lancet 1992;339: Russe SD, Koshkarian GM, Medinger AE, Carson PE, Higginbotham MB. Lack of effect of increased inspired oxygen concentrations on maxima exercise capacity or ventiation in stabe heart faiure. Am J Cardio 1999;84: Cark AL, Voterrani M, Swan JW, Coats AJ. The increased ventiatory response to exercise in chronic heart faiure: reation to pumonary pathoogy. Heart 1997;77: /hrt Cark AL, Coats AJ. Usefuness of arteria bood gas estimations during exercise in patients with chronic heart faiure. Br Heart J 1994;71: Odenburg O, Lamp B, Faber L, Tescher H, Horstkotte D, Topfer V. Seep-disordered breathing in patients with symptomatic heart faiure: a contemporary study of prevaence in and characteristics of 700 patients. Eur J Heart Fai 2007;9: Pembrey MS. Observations on Cheyne Stokes respiration. J Patho Bacterio 1908;12: Toyama T, Seki R, Kasama S, Isobe N, Sakurai S, Adachi H, et a. Effectiveness of nocturna home oxygen therapy to improve exercise capacity, cardiac function and cardiac sympathetic nerve activity in patients with chronic heart faiure and centra seep apnea. Circ J 2009;73: Sasayama S, Izumi T, Matsuzaki M, Matsumori A, Asanoi H, Momomura S, et a. Improvement of quaity of ife with nocturna oxygen therapy in heart faiure patients with centra seep apnea. Circ J 2009;73: NIHR Journas Library

121 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO Currow DC, Agar M, Smith J, Abernethy AP. Does paiative home oxygen improve dyspnoea? A consecutive cohort study. Paiat Med 2009;23: Uronis HE, Currow DC, McCrory DC, Samsa GP, Abernethy AP. Oxygen for reief of dyspnoea in midy- or non-hypoxaemic patients with cancer: a systematic review and meta-anaysis. Br J Cancer 2008;98: Cranston JM, Crockett A, Currow D. Oxygen therapy for dyspnoea in aduts. Cochrane Database Syst Rev 2008;3:CD Ben-Aharon I, Gafter-Gvii A, Pau M, Leibovici L, Stemmer SM. Interventions for aeviating cancer-reated dyspnea: a systematic review. J Cin Onco 2008;26: /JCO Booth S, Wade R, Johnson M, Kite S, Swannick M, Anderson H. The use of oxygen in the paiation of breathessness. A report of the expert working group of the Scientific Committee of the Association of Paiative Medicine. Respir Med 2004;98: j.rmed Abernethy AP, McDonad CF, Frith PA, Cark K, Herndon JE 2nd, Marceo J, et a. Effect of paiative oxygen versus room air in reief of breathessness in patients with refractory dyspnoea: a doube-bind, randomised controed tria. Lancet 2010;376: S (10) Guyatt GH, McKim DA, Austin P, Bryan R, Norgren J, Weaver B, et a. Appropriateness of domiciiary oxygen deivery. Chest 2000;118: Abernethy AP, Currow DC, Frith P, Fazekas BS. Prescribing paiative oxygen: a cinician survey of expected benefit and patterns of use. Paiat Med 2005;19: Currow DC, Christou T, Smith J, Carmody S, Lewin G, Aoun S, et a. Do terminay i peope who ive aone miss out on home oxygen treatment? An hypothesis generating study. J Paiat Med 2008;11: Chapman TH, Seymour J, Ho TBL. Home oxygen prescribing practices at a UK district genera hospita. Am J Respir Crit Care Med 2009;179:A ajrccm-conference _meetingabstracts.a Schwartzstein RM, Lahive K, Pope A, Weinberger SE, Weiss JW. Cod facia stimuation reduces breathessness induced in norma subjects. Am Rev Respir Dis 1987;136: /ajrccm/ Woheber AM, McKitrick DS, Davis SE. Designing research with hospice and paiative care popuations. Am J Hosp Paiat Care 2012;29: LeBanc TW, Lodato JE, Currow DC, Abernethy AP. Overcoming recruitment chaenges in paiative care cinica trias. J Onco Pract 2013;9: Jordhøy MS, Kaasa S, Fayers P, Ovreness T, Underand G, Ahner-Emqvist M. Chaenges in paiative care research; recruitment, attrition and adherence: experience from a randomized controed tria. Paiat Med 1999;13: Rector TS, Cohn JN. Assessment of patient outcome with the Minnesota Living with Heart questionnaire: reiabiity and vaidity during a randomized, doube bind, pacebo-controed tria of pimobendan. Am Heart J 1992;124: Rector TS, Kubo SH, Cohn JN. Vaidity of the Minnesota Living with Heart Faiure questionnaire as a measure of therapeutic response to enaapri or pacebo. Am J Cardio 1993;71: Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 95

122 REFERENCES 47. Witham MD, Crighton LJ, McMurdo MET. Using an individuaised quaity of ife measure in oder heart faiure patients. Int J Cardio 2007;116: Rector TS, Tschumperin LK, Kubo SH, Bank AJ, Francis GS, McDonad KM, et a. Use of the Living With Heart Faiure questionnaire to ascertain patients perspectives on improvement in quaity of ife versus risk of drug-induced death. J Card Fai 1995;1: (95)90025-X 49. Wyrwich KW, Tierney WM, Woinsky FD. Further evidence supporting an SEM-based criterion for identifying meaningfu intra-individua changes in heath-reated quaity of ife. J Cin Epidemio 1999;52: Abraham WT, Fisher WG, Smith AL, Deurgio DB, Leon AR, Loh E, et a. Cardiac resynchronization in chronic heart faiure. N Eng J Med 2002;346: Ceand JG, Cavert MJ, Verboven Y, Freemante N. Effects of cardiac resynchronization therapy on ong term quaity of ife: an anaysis from the Cardiac Resynchronisation-Heart Faiure (CARE-HF) study. Am Heart J 2009;157: Brooks R, EuroQoL Group. EuroQo: the current state of pay. Heath Poicy 1996;37: EuroQo Group. EQ-5D A Measure of Heath-Reated Quaity of Life Deveoped by the EuroQo Group. User Guide URL: (accessed 24 June 2015). 54. Wicock A, Crosby V, Carke D, Tattersfied A. Repeatabiity of breathessness measurements in cancer patients. Thorax 1999;54: Gift AG, Narsavage G. Vaidity of the numerica rating scae as a measure of dyspnoea. Am J Crit Care 1998;7: Oxberry SG, Band JM, Cark AL, Ceand JG, Johnson MJ. Minimay cinicay important difference in chronic breathessness: every itte heps. Am Heart J 2012;164: j.ahj Johnson MJ, Band JM, Abernethy A, Currow DC. Cinicay important differences in chronic refractory breathessness. J Pain Symptom Manage 2013;46: j.jpainsymman Haworth JE, Moniz-Cook E, Cark AL, Wang M, Waddington R, Ceand JG. Prevaence and predictors of anxiety and depression in a sampe of chronic heart faiure patients with eft ventricuar systoic dysfunction. Eur J Heart Fai 2005;7: Guyatt GH, Suivan MJ, Thompson PJ, Faen EL, Pugsey SO, Tayor DW, et a. The six minute wak: a new measure of exercise capacity in patients with chronic heart faiure. Can Med Assoc J 1985;132: Osson LG, Swedberg K, Cark AL, Witte KK, Ceand JG. Six minute corridor wak test as an outcome measure for the assessment of treatment in randomized, binded intervention trias of chronic heart faiure: a systematic review. Eur Heart J 2005;26: eurheartj/ehi Inge L, Sheton RJ, Rigby AS, Nabb S, Cark AL, Ceand JG. The reproducibiity and sensitivity of the 6-min wak test in edery patients with chronic heart faiure. Eur Heart J 2005;26: Schag CC, Heinrich RL, Ganz PA. Karnofsky performance status revisited: reiabiity, vaidity and guideines. J Cin Onco 1984;2: NIHR Journas Library

123 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO Mor V, Laiberte L, Morris JN, Wiemann M. The Karnofsky Performance Status Scae. An examination of its reiabiity and vaidity in a research setting. Cancer 1984;53: ( )53:9<2002::AID-CNCR >3.0.CO;2-W 64. Charson M, Szatrowski TP, Peterson J, God J. Vaidation of a combined comorbidity index. J Cin Epidemio 1994;47: Akaike H. A new ook at the statistica mode identification. IEEE Trans Automatic Contro 1974;19: Pickering RM, Weathera M. The anaysis of continuous outcomes in muti-centre trias with sma centre sizes. Stat Med 2007;26: Ariti CA, Ceand JGF, Pocock SJ, Pfeffer MA, Swedberg K, Granger CB, et a. Days aive and out of hospita and the patient journey in patients with heart faiure: Insights from the Candesartan in Heart faiure: Assessment of Reduction in Mortaity and morbidity (CHARM) program. Am Heart J 2011;162: Schuz KF, Atman DG, Moher D. CONSORT 2010 statement: updated guideines for reporting parae group randomized trias. Ann Intern Med 2010;152: Doan P. Modeing vauations for EuroQo heath states. Med Care 1997;35: Hayashi A, Tatsumi K, Kato K, Sakuma T, Okada O, Kimura H, et a. Compiance with ong term home oxygen therapy. Nihon Kyobu Shikkan Gakkai Zasshi 1996;34: Peckham DG, McGibbon K, Tonkinson J, Pimbey G, Pantin C. Improvement in patient adherence with ong term oxygen therapy foowing forma assessment with training. Respir Med 1998;92: Washaw MJ, Lim R, Evans CC, Hind CRK. Factors infuencing the adherence of patients using oxygen concentrators for ong term home oxygen therapy. Respir Med 1990;84: Evans TW, Waterhouse J, Howard P. Cinica experience with the oxygen concentrator. BMJ 1983;287: Diworth JP, Higgs CMB, Jones PA, White RJ. Acceptabiity of oxygen concentrators: the patients views. Br J Gen Pract 1990;40: Heaton RK, Grant I, McSweeny AJ, Adams KM, Petty TL. Psychoogic effects of continuous and nocturna oxygen therapy in hypoxemic chronic obstructive pumonary disease. Arch Intern Med 1983;143: Currow DC, Fazekas B, Abernethy AP. Oxygen use patients define symptomatic benefit discerningy. J Pain Symptom Manage 2007;34: j.jpainsymman Ritchie J, Spencer L. Quaitative Data Anaysis for Appied Poicy Research. In Bryman A, Burgess R, editors. Anayzing Quaitative Data. London: Routedge; pp _chapter_9 78. Camarri B, Eastwood PR, Cecins NM, Thompson PJ, Jenkins S. Six minute wak distance in heathy subjects aged years. Respir Med 2006;100: j.rmed Casanova C, Cei BR, Barria P, Casas A, Cote C, de Torres JP, et a. The 6-min wak distance in heathy subjects: reference standards from seven countries. Eur Respir J 2011;37: Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 97

124 REFERENCES 80. Inge L, Rigby AS, Carro S, Buttery R, King RF, Cooke CB, et a. Prognostic vaue of the 6 min wak test and sef-perceived symptom severity in oder patients with chronic heart faiure. Eur Heart J 2007;28: Suivan MJ, Higginbotham MB, Cobb FR. Increased exercise ventiation in patients with chronic heart faiure: intact ventiatory contro despite hemodynamic and pumonary abnormaities. Circuation 1988;77: Rubin SA, Brown HV, Swan HJ. Arteria oxygenation and arteria oxygen transport in chronic myocardia faiure at rest, during exercise and after hydraazine treatment. Circuation 1982;66: Munger MA, Stanek EJ, Nara AR, Stroh KP, Decker MJ, Nair RN. Arteria oxygen saturation in chronic congestive heart faiure. Am J Cardio 1994;73: (94) Herrin B, Syvén C. Increased arteria oxygen content an important compensatory mechanism in chronic moderate heart faiure. Cardiovasc Res 1991;25: Agostini PG, Wasserman K, Perego GB, Marenzi GC, Guazzi M, Assanei E, et a. Oxygen transport to musce during exercise in chronic congestive heart faiure secondary to idiopathic diated cardiomyopathy. Am J Cardio 1997;79: Vergeret J, Brambia C, Mounier L. Portabe oxygen therapy: use and benefit in hypoxaemic COPD patients on ong term oxygen therapy. Eur Respir J 1989;2: Howard P, Waterhouse JC, Biings CG. Compiance with ong term oxygen therapy by concentrator. Eur Respir J 1992;5: Restrick LJ, Pau EA, Braid GM, Cuinan P, Moore-Gion J, Wedzicha JA. Assessment and foow-up of patients prescribed ong term oxygen treatment. Thorax 1993;48: /thx Morrison D, Skwaraski K, MacNee W. Review of the prescription of domiciiary ong term oxygen therapy in Scotand. Thorax 1995;50: Granados A, Escarrabi J, Borras JM, Rodriguez-Roisin R. The importance of process variabes anaysis in assessment of ong term oxygen therapy by concentrator. Respir Med 1997;91: Ringbaek T, Lange P, Viskum K. Compiance with LTOT and consumption of mobie oxygen. Respir Med 1999;93: Atis S, Tutuogu B, Bugdayci R. Characteristics and adherence of patients receiving ong term oxygen therapy in Turkey. Monadi Arch Chest Dis 2001;56: Cark AL, Lammiman MJ, Goode K, Ceand JG. Is taking part in cinica trias good for your heath? A cohort study. Eur J Heart Fai 2009;11: Semigran MJ, Cockri BA, Kacmarek R, Thompson BT, Zapo WM, Dec GW, et a. Hemodynamic effects of inhaed nitric oxide in heart faiure. J Am Co Cardio 1994;24: / (94) Mahajan A, Shabanie A, Varshney SM, Marijic J, Sopher MJ. Inhaed nitric oxide in the preoperative evauation of pumonary hypertension in heart transpant candidates. J Cardiothorac Vasc Anesth 2007;21: Leventha H, Brissette I, Leventha EA. The Common-Sense Mode of Sef-Reguation of Heath and Iness. In Cameron LD, Leventha H, editors. The Sef-Reguation of Heath and Iness Behaviour. London: Routedge; pp NIHR Journas Library

125 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO Burish TG, Bradey LA, Nerenz DR, Leventha H. Sef-Reguation Theory in Chronic Iness. In Burish TG, Bradey LA, editors. Coping with Chronic Disease: Research and Appications. New York, NY: Academic Press; pp Hagger MS, Orbe S. A meta-anaytic review of the common-sense mode of iness representations. Psycho Heath 2003;18: Green CP, Porter CB, Bresnahan DR, Spertus JA. Deveopment and evauation of the Kansas City Cardiomyopathy Questionnaire: a new heath status measure for heart faiure. J Am Co Cardio 2000;35: Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 99

126

127 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Appendix 1 Origina Heath Technoogy Assessment commissioning brief NHS R&D Heath Technoogy Assessment programme HTA number 06/80 Home oxygen therapy for chronic heart faiure Introduction The aim of the HTA programme is to ensure that high quaity research information on the costs, effectiveness and broader impact of heath technoogies is produced in the most efficient way for those who use, manage, provide care in or deveop poicy for the NHS. Topics for research are identified and prioritised to meet the needs of the NHS. Heath technoogy assessment forms the argest portfoio of work in the NHS Research and Deveopment Programme and each year about fifty new studies are commissioned to hep answer questions of direct importance to the NHS. The studies incude both primary research and evidence synthesis. Question What is the cinica and cost-effectiveness of home oxygen therapy in addition to standard care for patients with chronic heart faiure? 1 Technoogy Home oxygen therapy pus standard care. 2 Patient group Patients with chronic heart faiure and chronic arteria hypoxia. Further incusion criteria to be estabished in Phase I of study. 3 Setting Home. 4 Contro or comparator treatment Standard care. 5 Design Two stage Stage 1: a feasibiity study: An assessment of the practica and ethica issues invoved in carrying out a RCT, and the determination of the most appropriate incusion criteria for a RCT incuding at east a definition of hypoxia in terms of PaO 2 and of heart faiure in terms of ejection fraction (if researchers fee other definitions are more cinicay appropriate they shoud justify their decision), the most appropriate oxygen fow rate and timing, and the most appropriate mechanism of deivery. Stage 2: a randomised controed tria to assess the use of home oxygen therapy in addition to standard care for patients with heart faiure. Resuts for patients with different grades of chronic heart faiure wi be modeed. 6 Primary outcomes Stage 1 The appicants shoud propose appropriate objective success criteria to move onto stage 2. These might incude approva from ethics committees, centres signed up to participate, demonstration of equipoise among cinicians ikey to enter patients into a tria and the abiity to recruit adequate numbers of patients to the tria. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 101

128 APPENDIX 1 Stage 2 Mortaity, QoL and quaity-adjusted ife-years (QALYs) gained. Secondary outcomes: breathessness, drowsiness, mortaity, adverse effects, functiona capacity, cost-effectiveness. 7 Minimum duration of foow-up Two years. 8 Note to appicants Pease submit separate costings for Stage 1 and Stage 2 in the outine appication: the costs shoud be detaied in the Summary of Project section on page 2 of the eectronic appication form. The combined cost of both stages shoud be entered in the Research grant box on page 1 of the form. Stage 2 wi ony proceed if the pre-agreed success criteria for stage 1 are met. If fu proposas are requested, appicants wi need to submit protocos for both stages of the study. In the event of a sma gap between stage 1 and stage 2 funding may be avaiabe to ensure continuity. Background to commissioning brief Patients with chronic heart faiure may deveop arteria hypoxaemia, which can have acute and chronic adverse effects on cardiac function. Patients with grade III(b) or IV heart faiure and chronic arteria hypoxaemia have particuar disturbance in breathing at night, which causes disturbed seep, daytime seepiness and reduced abiity to function. Up to 50% die within 1 year. Oxygen therapy may be provided for continuous use by patients with chronic hypoxaemia at home. However, current recommendations by the Roya Coege of Physicians are not based on cear evidence of effectiveness or cost-effectiveness. There is a ack of evidence for oxygen therapy in patients with chronic heart faiure. Notes to appicants For many of the questions posed by the HTA programme, a randomised controed tria is ikey to be the most appropriate method of providing an answer. However, there may be practica or ethica reasons why this might not be possibe. Appicants proposing other research methods are invited to justify these choices. Appicants are asked to: 1. Foow the Medica Research Counci s Good Cinica Practice guideines ( pdf) when panning how studies, particuary RCTs, wi be supervised. Further advice specific to each topic wi be given by the HTA programme at fu proposa and contract stages. 2. Note that trias invoving medicina products must compy with The Medicines for Human Use (Cinica Trias) Reguations In the case of such trias, the Department of Heath (DH) expects the empoying institution of the chief investigator to be nominated as the sponsor. Other institutions may wish to take on this responsibiity or agree co-sponsorship with the empoying institution. The DH is prepared to accept the nomination of mutipe sponsors. Appicants who are asked to submit a fu proposa wi need to obtain confirmation of a sponsor(s) to compete their appication. The DH reserve the right to withdraw from funding the project if they are not satisfied with the arrangements put in pace to conduct the tria. The MHRA ([email protected], can provide guidance whether your tria woud be covered by the reguations. The DH/MRC website ( aso contains the atest information about Cinica Trias reguations and a hepfu FAQ page. 102 NIHR Journas Library

129 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Making an appication If you wish to submit an outine proposa on this topic, compete the eectronic appication form and return it to the HTA Commissioning Manager at the Nationa Coordinating Centre for Heath Technoogy Assessment, Maipoint 728 Bodrewood, University of Southampton, Southampton SO16 7PX by 7 February Outine appications wi be considered by the HTA Commissioning Board at its meeting in Juy If they are acceptabe, investigators wi be given a minimum of eight weeks to submit a fu proposa. Appications received after 1300 hours on the due date wi not be considered. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 103

130

131 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Appendix 2 Reguatory approvas and detais of study sites TABLE 41 Timing of MREC approvas MREC approvas Origina ethics approva, 24 August 2009 Substantia amendment 1, 11 Apri 2011 Substantia amendment 2, 20 February 2013 Substantia amendment 3, 21 March 2013 Substantia amendment 4, 2 August 2013 Substantia amendment 5, 26 March 2014 Change Approva for the NEON study Change in study design to an open three-arm tria Change in study design from three-arm to two-arm tria (BMT ony vs. LTOT pus BMT). Primary outcome time point was changed from 12 months to 6 months Permission to use invitation etter for potentia participants from patient ists hed at NHS genera practitioner practices (patient identification centres), or via existing ists of ikey eigibe patients hed within NHS Hospitas Permission for researchers and study nurses to be abe to contact patients via phone after initia introductory etter from the principa investigator New patient etter informing them of end of tria TABLE 42 Timings of approvas for sites Research site Principa investigator Research and design approva County Durham & Darington NHS Foundation Trust (Darington Site) Professor Jerry Murphy 22 August 2011 Hu and East Yorkshire NHS Trust Professor Andrew Cark 24 February 2012 The Mid Yorkshire Hospitas NHS Trust Dr Pau Brooksby 11 Apri 2012 NHS Tayside Dr Mies Witham 15 June 2012 University Hospitas of Leicester Professor Iain Squire 9 Juy 2012 Chesterfied Roya Hospita Dr Justin Cooke 5 Juy 2012 Barnet and Chase Farm Hospitas NHS Trust Dr Ameet Bakhai 2 August 2012 Eaing Hospita NHS Trust Dr Stuart Rosen 6 September 2012 County Durham & Darington NHS Foundation Trust (Durham site) Dr Mohamed E-Harari 25 September 2012 Penine Acute Hospitas NHS Trust a Dr Joanta Soboewska 19 October 2012 NHS Bradford and Airedae Dr Pau Smith 15 November 2012 Aneurin Bevan University Heath Board Dr Jackie Austin 9 January 2013 City Hospitas Sunderand NHS Foundation Trust Dr John Baxter 21 March 2013 Pymouth Hospitas NHS Trust Dr Andrew Stone 3 June 2013 East Cheshire NHS Trust Dr Robin Egde 13 December 2014 a For the Penine Acute Hospitas NHS Trust, Odham and Bury Primary Care Trusts ony were approved on 19 October 2012; approva was reissued on 12 November 2012 to incude Rochdae, Heywood and Middeton). Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 105

132

133 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Appendix 3 Home oxygen therapy patient information sheet version 7 The HOT Study Does Home Oxygen Therapy (HOT) in Addition to Standard Care Reduce Disease Severity and Improve Symptoms in Patients with Chronic Heart Faiure? We woud ike to invite you to take part in a research study. Before you decide, you need to understand why the research is being done and what it woud invove for you. Pease take time to read the foowing information carefuy. Tak to others about the study if you wish. Part 1 tes you the purpose of the study and what wi happen to you if you take part. Part 2 gives you more detaied information about the conduct of the study. Ask us if there is anything that is not cear or if you woud ike more information. Take time to decide whether or not you wish to take part. Part 1 1. What is the purpose of the study? Home Oxygen Therapy can be given in a patient s own home by instaing an oxygen concentrator machine. This fiters ordinary room air to concentrate the oxygen, and then deivers it through a narrow tube to the patient. Therapy can be deivered overnight and for severa hours during the daytime, too at east 15 hours out of the 24 (ong term oxygen therapy). Home Oxygen Therapy is a proven treatment to hep peope with ongstanding ung probems and ow oxygen bood eves. Sometimes it is used to try and hep the breathessness that can be caused by heart faiure. However, we do not know whether it does actuay hep patients with heart faiure or not, or, if it does, whether night time or ong term oxygen is more effective. As the treatment itsef is quite expensive, and can cause a burden because of the equipment needed, we fee it is important to ask these questions: is it hepfu, do patients find it a burden, and how shoud we give it? We are testing to see whether having extra oxygen deivered by the concentrator makes a difference to your symptoms, bood oxygen eves and degree of heart faiure. In this study, we wi compare patients who do not receive Home Oxygen Therapy to those who receive it for 15 hours during the day and overnight. We intend to invite 222 patients from different hospitas to participate in this study. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 107

134 APPENDIX 3 2. Why have I been chosen? We have invited you to participate in our study because you have been identified as having heart faiure causing symptoms such as breathessness. 3. Do I have to take part? No. It is entirey up to you to decide. We wi discuss the study and go through this information sheet, which we wi then give to you. After you have had time to decide about taking part, we wi then ask you to sign a consent form to show you have agreed. You are free to withdraw at any time, without giving a reason. Shoud you wish to withdraw you shoud et your oca cinica team know or write/ca the York Cinica Trias Unit (see page 9). This woud not affect your usua care, or effect your reationship with your doctor or the hospita staff. The ony information that we wi keep for research and anaysis is what has been coected about you unti the time you decide to stop participating in the tria. 4. What are the aternatives for treatment? You are aready receiving the optima medication known to be effective for your condition. This study is ooking at an additiona treatment, rather than an aternative one. 5. What wi happen to me if I take part? After you sign the consent form, you wi be randomised (a process by which a computer determines by chance which treatment you receive) to receive ong term oxygen therapy for 15 hours a day or no oxygen therapy. You woud remain on this treatment for at east six months. Some peope may stay on the study for onger, up to a maximum of 2 years, depending on when they start the tria. We wi te you how ong we expect you to stay in the tria before you agree to join the tria. The study wi not interfere with your current medication. You can withdraw from the study at any time, without giving a reason. Start of the study We wi invite you to attend the cinic at <insert hospita name>. At the hospita, we need to measure your degree of heart faiure, and how much troube it causes you at the start of the study so we can see whether the home oxygen therapy makes any difference. We wi do this by asking you to fi out some questionnaires and to undergo some tests described beow many of which you may aready have had during your treatment for heart faiure: Puse, bood pressure and breathing rate A wak test (in which you wi wak up and down a corridor within the cinic for up to 6 minutes) An oxygen bood eve measurement (done with a simpe monitor that cips onto your finger and reads the oxygen eve in the bood through your skin this takes about a minute) to observe the eve of oxygen in your bood wi be performed before and after the 6 minute wak test A bood test to measure a substance in the bood that gives an indication of the degree of heart faiure (caed NT-proBNP) 108 NIHR Journas Library

135 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 An echocardiogram (to obtain an utra-sound picture of your heart) Questionnaires that ask about your symptoms, other inesses, how you manage day to day and how much your heart faiure affects your ife and your mood. One of the questionnaires wi aso ask you about how we you seep We envisage that each visit wi ast minutes. You may be asked to compete an overnight seep test. This is a simpe painess test that measures your breathing pattern and the eves of oxygen in your bood overnight. We show you how to use the machine at home, and we wi coect the machine after you have competed the test. Your consutant or research nurse wi advise you whether you wi have a seep study or not. [This section to be deeted at sites not undertaking the seep study.] If you are randomised to receive a home oxygen concentrator, we wi arrange for it to be deivered and instaed in your home. You wi be shown how to use it, and be given a eafet about the machine (attached). You can contact the research nurse, the company, or your cinician in the event of any probem or query (pease see page 9 for contact detais). The NHS wi pay for the eectricity costs of the machine (attached). At 3 months You wi see the study nurse again at 3 months for the foowing: Puse, bood pressure and breathing rate A bood test Questionnaires that ask about your symptoms and how you manage day to day. We envisage that this visit wi ast minutes. If you don t need to be at the hospita, and prefer to be seen at home or a oca cinic, then the nurse wi visit you there. At 6 months We wi repeat a the tests that we did at the start, except for the echocardiogram. This wi be done aongside your norma hospita appointment at <insert hospita name>. We envisage that each visit wi ast minutes. At 12 months We wi repeat a the tests that we did at the start. This wi be done aongside your norma hospita appointment at < hospita name>. We envisage that each visit wi ast minutes. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 109

136 APPENDIX 3 If you stay in the tria onger than a year, then you woud have another assessment at 18 months (the same as the 6 months visit), and possiby an assessment at 24 months (the same as the 12 months visit). Both of these visits woud happen aongside your routine cinic visits to the <insert hospita name>. When you have competed the tria, we wi arrange for remova of the oxygen concentrator, if that is what you were randomised to receive. If you have found the oxygen concentrator beneficia, and wish to continue treatment with it, we wi eave it in pace. 6. What wi I have to do? There are no drugs invoved in this study. Norma room air contains 21% oxygen. The concentrator devices deiver increased oxygen (28%) to the nostris through a narrow tube. We are testing to see whether having extra oxygen deivered by the concentrator makes a difference to your symptoms, bood oxygen eves and degree of heart faiure. You wi attend your norma 6 monthy cinics at <insert hospita name>, and wi undergo some additiona tests aongside your norma tests, and compete some questionnaires (these are described above). It may be more comfortabe for you to attend in oose fitting cothes with comfortabe waking shoes for the 6 minute wak test. We wi do one additiona assessment when you have been in the tria for 3 months. This woud normay occur in your home, with a nurse visiting you, but may be in hospita or a oca cinic, depending on your needs and preferences. This assessment comprises of some minor tests and questionnaires (these are described above). If you are randomised to receive an oxygen concentrator, we wi arrange for it to be instaed in your home after you start the tria, and uninstaed when you compete the tria. If you have found the home oxygen therapy usefu, you can continue to use the oxygen concentrator. 7. What does having an Oxygen Concentrator invove? A concentrator machine concentrates the oxygen from norma room air to provide a continuous suppy of higher concentration of oxygen than norma. A fuy trained homeoxygen engineer wi deiver the concentrator machine to your home, insta it and provide you with a the necessary tubing. The engineer wi instruct you in its use and you wi be given instructions for simpe weeky maintenance. Modern concentrators are reiabe, quiet, easy-to-use and most patients find them a convenient source of oxygen. The engineer wi service the machine after 6 months of use. At the end of the study, the engineer wi remove the concentrator. Any excess in your eectricity bis due to the concentrator wi be reimbursed. During the study, an engineer wi be avaiabe in office hours if there are probems with the machine and can be contacted through your study nurse <insert name and teephone number>. As there is no proven need for you to have oxygen, there wi be no need for emergency repair and it wi be safe for you to wait for office hours. 110 NIHR Journas Library

137 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 You are aready receiving the optima medication known to be effective for your condition. This study is ooking at an additiona treatment, rather than an aternative one. If your medica condition shoud change such that it is thought that you do need oxygen therapy, that wi be provided by your doctors in the usua manner. 8. Expenses and payments We wi pay for any reasonabe trave expenses, and any excess eectricity bis arising from the use of the oxygen concentrator. We wi arrange and pay for the coection of overnight testing equipment from your home. 9. What are the possibe disadvantages and risks of taking part? The oxygen devices wi take up a sma amount of space and make a ow humming noise which some peope may find intrusive. Compying with safety measures wi be important, particuary with regard to smoking. Bood samping may cause a sma amount of beeding, discomfort, or a bruise, and in very rare cases, infection. Occasionay a person may fee ight headed when their bood is drawn. 10. What are the side effects of any treatment received when taking part? In addition to the safety issues on page 7, you may experience some drying of nasa passages and pressure of tubing over tops of ears. Both can usuay be heped with simpe measures. 11. What are the possibe benefits of taking part? We cannot promise the study wi hep you but the information we get from this study wi hep improve our understanding of the effects oxygen therapy in peope with heart faiure. We hope the information wi aow us to answer the questions about oxygen therapy posed earier, so we can recommend appropriate use for patients with heart faiure throughout the NHS. 12. What happens when the research study stops? If you fee a benefit from the home oxygen therapy, then you can continue to use it when you compete the tria. Otherwise, your routine care wi continue as before. Of course, if your condition shoud change during the study, such that you definitey need oxygen therapy anyway at the end of the study, then that wi be provided as part of your usua management even when the study has ended. 13. What wi happen if I don t want to carry on with the study? You are free to withdraw at any time for any reason. This wi not affect your future care. If you withdraw from the study and you state that you do not want the information we Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 111

138 APPENDIX 3 have coected about used in this study we wi destroy a of the data we have coected about you. 14. What if there is a probem? If you have a concern about any aspect of this study, pease speak with the study nurse who wi do their best to answer a your questions. You are wecome to discuss your concerns with other members of the cinica team. You may wish to contact the Patient Advice and Liaison Service. These contact detais are isted on page 9. If you wish to compain, or have any concerns about any aspect of the way you have been approached or treated during this study, the norma NHS compaints mechanisms are avaiabe to you. Information about patient rights, research-reated questions and research-reated injury can be obtained from the Loca Patients Advice and Liaison Service or the British Heart Foundation. 15. What if something goes wrong? In the unikey event that something does go wrong and you are harmed, or fee you are harmed, during the research and this is due to someone s negigence then you may have grounds for a ega action against Hu and East Yorkshire NHS Trust. 16. Wi my taking part in the study be kept confidentia? Yes. A information coected about you during the course of the study wi be kept stricty confidentia and in accordance with The Data Protection Act 1998 and no information by which patients can be identified wi be reproduced or discosed. Your GP and his/her Genera Practice wi be notified of your participation in the study, with your permission. The persona data recorded on a records wi be regarded as confidentia, and to preserve each patient s anonymity, ony your initias and date of birth wi be recorded on the forms associated with the study. Patients wi be identified by the use of a unique tria number aocated to them on entry into the study. The study doctor wi use your persona data for the purposes of administering and conducting the study and wi ensure that strict patient confidentiaity is maintained. At a minimum, this data woud incude name, date of birth and reevant NHS patient identifiers as required. The data wi aso be accessibe to approved members of the Cinica Trias Team incuding the University of York Trias Unit, and reguatory authorities for approved tria purposes ony. The NHS approved oxygen suppy company wi provide the tria with readings from the home oxygen equipment. The data wi be anaysed in accordance with the European Union Directive on the protection of individuas with regard to the processing of persona data. 17. Contact Detais: Principa Loca Investigator: Research Nurse: Regiona oxygen suppier Emergency contact: Patient Advice & Liaison Service British Heart Foundation: York Cinica Trias Unit: Mrs Sarah Cockayne 112 NIHR Journas Library

139 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 If you wish to contact the research team by post for any reason, pease write to: Mrs Sarah Cockayne, Research Feow XXXX Te: XXXX This competes Part 1 of the Information Sheet. If the information in Part 1 has interested you and you are considering participation, pease continue to read the additiona information in Part 2 before making any decision. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 113

140 APPENDIX 3 Part 2 1. What if new information becomes avaiabe? Sometimes we get new information about the treatment being studied. If this happens, your research doctor wi te you and discuss whether you shoud continue in the study. If you and/or your research doctor decide you shoud not carry on, your research doctor wi make arrangements for your care to continue. If your research doctor is happy for you to continue in the study, and you agree, he may ask you to sign an updated consent form. If the study is stopped for any other reason, we wi te you and arrange your continuing care. 2. What wi happen to any sampes I give? Bood sampes wi be anonymised, stored, anaysed and destroyed by your oca hospita aboratory If they are not anaysed for any reason, they wi be destroyed. The anonymised anaysis data wi be passed to the University of York Cinica Tria Unit, who are overseeing data management for this tria. 3. Wi any genetic tests be done? No. 4. What about my Genera Practitioner? We wi inform your GP (with your permission) that you have been invited to participate in the study. 5. What wi happen to the resuts of the research study? These wi be examined by the staff undertaking the study and pubished as appropriate in medica journas, for the benefit of other patients and research studies. You wi not be personay identified in any report/ pubication. You may request resuts of the tria if you so wish. An executive summary wi be prepared for patient heart faiure support groups to see on request. 6. Who is organising and funding the research? The research is funded by the Heath Technoogy Assessment (HTA) programme, which is part of the Department of Heath. The sponsor for the tria is Hu and East Yorkshire NHS Trust. The research group is being ed by Professor Andrew Cark from the Department of Academic Cardioogy, Caste Hi Hospita, Hu. Data coection is being managed by the University of York Cinica Trias Unit. Both Hu and East Yorkshire NHS Trust and the University of York Cinica Trias Units are part of the Hu-York Medica Schoo (HYMS). 114 NIHR Journas Library

141 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO Who has reviewed the study? The Northern and Yorkshire REC has reviewed this study for adherence with medica and ethica standards and scientific vaue and has given a favourabe ethica opinion for conduct in the NHS. Thank you for reading this information. If you decide to take part in the tria you wi be given a copy of this information sheet and signed consent form to keep. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 115

142

143 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Appendix 4 Patient consent form Tite of Study: Hot Tria - Does home oxygen therapy (HOT) in addition to standard care reduce disease severity and improve symptoms in patients chronic heart faiure? Pease initia box 1. I confirm that I have read and understand the information sheet dated 20/02/2013 (version 7) for the above study and have had the opportunity to ask questions. 2. I understand that my participation is vountary and that I am free to withdraw at any time, without giving any reason, without my medica care or ega rights being affected. 3. I understand that sections of any of my medica notes may be ooked at by responsibe individuas from the sponsor (Hu and East Yorkshire Hospitas NHS Trust) the research staff or from reguatory authorities where it is reevant to my taking part in research; I give permission for these individuas to have access to my records. 4. I agree that my Genera Practitioner wi be informed of my participation in this study and wi be advised of any significant information reating to my heath that comes to ight. 5. I consent to the transfer, storage and use of paper and eectronic persona information, for the purposes of this study by the research team incuding the University of York. I understand that any information that coud identify me wi be kept stricty confidentia and that no persona information wi be incuded in the study report or other pubication. 6. I consent to donating bood sampes that wi be stored, managed and processed anonymousy/unnamed, and destroyed at the end of the tria or before. 7. I understand that a the information coected on my behaf wi be kept stricty confidentia and treated according to the European Union Directive (DIR95/46/EC) on the protection of individuas with regard to the processing of persona data and on the free movement of such data and oca appicabe aws. 8. I give my consent to take part in the HOT tria. Name of Patient Date Signature Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 117

144

145 DOI: /hta19750 HEALTH TECHNOLOGY ASSESSMENT 2015 VOL. 19 NO. 75 Appendix 5 Graphica checks of the assumptions for the primary anaysis mode Density Standardised residuas FIGURE 11 Histogram of the standardised residuas. Queen s Printer and Controer of HMSO This work was produced by Cark et a. under the terms of a commissioning contract issued by the Secretary of State for Heath. This issue may be freey reproduced for the purposes of private research and study and extracts (or indeed, the fu report) may be incuded in professiona journas provided that suitabe acknowedgement is made and the reproduction is not associated with any form of advertising. Appications for commercia reproduction shoud be addressed to: NIHR Journas Library, Nationa Institute for Heath Research, Evauation, Trias and Studies Coordinating Centre, Apha House, University of Southampton Science Park, Southampton SO16 7NS, UK. 119

146 APPENDIX 5 2 Standardised residuas Inverse norma 2 FIGURE 12 Q Q pot of the standardised residuas. 2 1 Standardised residuas FIGURE 13 Scatterpot of residuas vs. predicted vaues. Linear prediction, fixed portion 120 NIHR Journas Library

147

148 EME HS&DR HTA PGfAR PHR Part of the NIHR Journas Library This report presents independent research funded by the Nationa Institute for Heath Research (NIHR). The views expressed are those of the author(s) and not necessariy those of the NHS, the NIHR or the Department of Heath Pubished by the NIHR Journas Library