BMJ Open Is Committed to Open Peer Review. As Part of This Commitment We Make the Peer Review History of Every Article We Publish Publicly Available

BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from

BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.

When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.

The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.

BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay-per-view fees (http://bmjopen.bmj.com).

If you have any questions on BMJ Open’s open peer review process please email [email protected] http://bmjopen.bmj.com/ on October 1, 2021 by guest. Protected copyright. BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from

Processes of care and survival associated with treatment in specialist teenage and young adult cancer centres: results from the BRIGHTLIGHT cohort study ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2020-044854

Article Type: Original research

Date Submitted by the 16-Sep-2020 Author:

Complete List of Authors: Fern, Lorna; University College London Hospitals NHS Foundation Trust, Oncology Taylor, Rachel; University College London Hospitals NHS Foundation Trust, Centre for Nurse, Midwife and AHP Led Research (CNMAR) Barber, Julie; University College London, Department of Statistical Science Alvarez-Galvez, Javier; University of Cadiz, Department of Biomedicine, Biotechnology and Public Health Feltbower, Richard; University of Leeds, Lea, Sarah; University College London Hospitals NHS Foundation Trust Martins, Ana; University College London Hospitals NHS Foundation Trust, Cancer Clinical Trials Morris, Stephen; University of Cambridge, Primary Care Unit http://bmjopen.bmj.com/ Hooker, Louise; University Hospital Southampton NHS Foundation Trust, Wessex Teenage and Young Adult Cancer Service, Gibson, Faith; University of Surrey, Faculty of Health and Medical Sciences; Great Ormond Street Hospital For Children NHS Foundation Trust, Raine, Rosalind; University College London, Institute of Epidemiology & Health Stark, Dan; University of Leeds, Leeds Insitute of Molecular Medicine

Whelan, Jeremy; University College London Hospitals NHS Foundation on October 1, 2021 by guest. Protected copyright. Trust,

Organisational development < HEALTH SERVICES ADMINISTRATION & Keywords: MANAGEMENT, Quality in health care < HEALTH SERVICES ADMINISTRATION & MANAGEMENT, ONCOLOGY

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 25 BMJ Open

1 2 3 4 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 7 8 9 I, the Submitting Author has the right to grant and does grant on behalf of all authors of the Work (as defined 10 in the below author licence), an exclusive licence and/or a non-exclusive licence for contributions from authors 11 who are: i) UK Crown employees; ii) where BMJ has agreed a CC-BY licence shall apply, and/or iii) in accordance 12 with the terms applicable for US Federal Government officers or employees acting as part of their official 13 duties; on a worldwide, perpetual, irrevocable, royalty-free basis to BMJ Publishing Group Ltd (“BMJ”) its 14 licensees and where the relevant Journal is co-owned by BMJ to the co-owners of the Journal, to publish the 15 Work in this journal and any other BMJ products and to exploit all rights, as set out in our licence. 16 17 The Submitting Author accepts and understands that any supply made under these terms is made by BMJ to 18 the Submitting Author Forunless you peer are acting as review an employee on behalf only of your employer or a postgraduate 19 student of an affiliated institution which is paying any applicable article publishing charge (“APC”) for Open 20 Access articles. Where the Submitting Author wishes to make the Work available on an Open Access basis (and 21 intends to pay the relevant APC), the terms of reuse of such Open Access shall be governed by a Creative 22 Commons licence – details of these licences and which Creative Commons licence will apply to this Work are set 23 out in our licence referred to above. 24 25 Other than as permitted in any relevant BMJ Author’s Self Archiving Policies, I confirm this Work has not been 26 accepted for publication elsewhere, is not being considered for publication elsewhere and does not duplicate 27 material already published. I confirm all authors consent to publication of this Work and authorise the granting 28 of this licence. 29 30 31 32 33 34 35 36

37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44

45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 25

1 2 3 Processes of care and survival associated with treatment in specialist teenage and young 4 adult cancer centres: results from the BRIGHTLIGHT cohort study 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 7 Lorna A Fern PhD, Rachel M Taylor PhD, Julie A Barber PhD, Javier Alvarez Galvez PhD, 8 Richard Feltbower PhD, Sarah Lea PhD, Ana Martins PhD, Steve Morris PhD, Louise Hooker 9 MSc, Faith Gibson PhD, Rosalind Raine PhD, Dan P Stark MD, Jeremy S Whelan MD 10 11 12 13 Lorna Fern, Teenage and Young Adult Cancer Researcher and Patient/Public Involvement 14 Lead, Cancer Division, University College London Hospitals NHS Foundation Trust, London 15 NW1 2PG 16 17 18 Rachel Taylor, DirectorFor of the peer CNMAR, Centrereview for Nurse, Midwifeonly and AHP Led Research 19 (CNMAR), University College London Hospitals NHS Foundation Trust, London NW1 2PG 20 21 Julie Barber, Associate Professor in Medical Statistics, Department of Statistical Science, 22 23 University College, London WC1E 6BT 24 25 Javier Alvarez-Galvez, Research Fellow, Department of Biomedicine, Biotechnology and 26 Public Health, University of Cadiz, Avda. Ana de Viya, 52, 11009 Cádiz, Spain 27 28 Richard Feltbower, Senior Lecturer in Epidemiology, Leeds Institute for Data Analytics, 29 30 School of Medicine, University of Leeds, Leeds, LS2 9JT. 31 32 Sarah Lea, Research Facilitator, Cancer Division, University College London Hospitals NHS 33 Foundation Trust, London NW1 2PG 34 35 36 Ana Martins, Research Associate, Cancer Division, University College London Hospitals

37 NHS Foundation Trust, London NW1 2PG http://bmjopen.bmj.com/ 38 39 Steve Morris, Professor of Health Economics, Primary Care Unit, Department of Public 40 41 Health and Primary Care, University of Cambridge, United Kingdom 42 43 Louise Hooker, teenage Cancer Trust Lead Nurse, Wessex Teenage and Young Adult 44 Cancer Service, University Hospital Southampton, Southampton, SO16 5YA 45 on October 1, 2021 by guest. Protected copyright. 46 47 Faith Gibson, Professor of Child Health and Cancer Care, School of Health Sciences, 48 Faculty of Health and Medical Sciences, University of Surrey, Guildford, Surrey GU2 7XH 49 and Centre for Outcomes and Experience Research in Children’s Health, Illness and 50 Disability (ORCHID), Great Ormond Street Hospital for Children NHS Foundation Trust, 51 London WCIN3JH 52 53 54 Rosalind Raine, Professor of Health Services research, Department of Applied Health 55 Research, Director, NIHR CLAHRC North Thames, NIHR Senior Investigator, University 56 College London, London WC1E 7HB 57 58 59 Dan Stark, Professor of Teenage and Young Adult Cancer Research and Consultant 60 Medical Oncologist, Leeds Institute of Medical Research at St James's, Leeds LS9 7TF

1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 25 BMJ Open

1 2 3 4 Jeremy Whelan, Professor of Cancer Medicine, Cancer Division, University College London 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 Hospitals NHS Foundation Trust, London NW1 2PG 7 8 Correspondence: Dr Rachel Taylor, [email protected] 9 10 Keywords: BRIGHTLIGHT, teenagers and young adults, cancer, observational research, 11 12 cohort, outcome, survival 13 14 Abstract = 297/300 15 Word count = 3,536 16 17 Number of Tables = 6 18 Number of Figures For= 1 peer review only 19 20 21 ABSTRACT 22 23 24 Objective: Survival gains in teenagers and young adults (TYA) are reported to be lower than 25 children and adults for some cancers. Place of care is implicated, influencing access to 26 specialist TYA professionals and research. Consequently, age-appropriate specialist cancer 27 28 care is advocated for TYA although systematic investigation of associated outcomes is 29 lacking. In England, age-appropriate care is delivered through 13 Principal Treatment 30 Centres (TYA-PTC). BRIGHTLIGHT is the national evaluation of TYA cancer services to 31 examine outcomes associated with differing places and levels of care. We aimed to examine 32 the association between exposure to TYA-PTC care, survival and documentation of clinical 33 34 processes of care. 35 36 Design: Prospective cohort study 37 http://bmjopen.bmj.com/ 38 Setting: 109 National Health Service (NHS) hospitals across England 39 40 41 Participants: 1,114 TYA, aged 13-24, newly diagnosed with cancer between 2012-2014 42 43 Intervention: Participants were assigned a TYA-PTC category dependent on the proportion 44

45 of care delivered in a TYA-PTC in the first year after diagnosis: all care in a TYA-PTC (ALL- on October 1, 2021 by guest. Protected copyright. 46 TYA-PTC, n=270), no care in a TYA-PTC (NO-TYA-PTC, n=359), and some care in a TYA- 47 PTC with additional care in a children’s/adult unit (SOME-TYA-PTC, n=419). 48 49 Primary outcome: Data were collected on documented processes indicative of age- 50 51 appropriate care using clinical report forms, and survival through linkage to NHS databases. 52 53 Results: TYA receiving NO-TYA-PTC care were less likely to have documentation of 54 molecular diagnosis, be reviewed by a children’s or TYA multi-disciplinary team, be 55 assessed by supportive care services or have a fertility discussion. There was no significant 56 57 difference in survival according to category of care. There was weak evidence that the 58 association between care category and survival differed by age (p=0.08) with higher hazard 59 ratios for those over 19 receiving ALL or SOME-TYA-PTC compared with NO-TYA-PTC. 60

2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 25

1 2 3 Conclusion: TYA-PTC care was associated with better documentation of clinical processes 4 associated with age-appropriate care but not improved survival. 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 7 8 Strengths & limitations of this study 9  Our study is the first prospective longitudinal national evaluation of specialist cancer 10 11 care for teenagers and young adults with cancer. 12  Using routinely collected National Health Service (NHS) data we were able to assign 13 participants into three groups according to how much exposure to specialist care they 14 had received in the first year following diagnosis. 15 16  Multiple data sources from patients, NHS and clinical report forms allowed us to 17 adjust for multiple predefined confounding variables. 18  Specialist servicesFor for peer teenagers reviewand young adults only have evolved since recruitment 19 and may not reflect current service configuration. 20  Our study did not meet its anticipated recruitment target, recruiting 20% of the total 21 22 population diagnosed during the recruitment period and this may limit generalisability 23 of the results. 24 25 26 27 INTRODUCTION 28 29 Cancer in teenagers and young adults (TYA) aged 15-24 years is rare, comprising 30 approximately 1% of the total cancer population in the United Kingdom (UK)(1). Historically 31 when compared to children and older adults, TYA with cancer have experienced lower 32 33 survival improvements for certain cancers. Prolonged pathways to cancer diagnosis, less 34 research, an inadequate understanding of cancer biology in young people, poor choice of 35 treatment protocols and place of care are all implicated (2-7). It is now generally accepted 36 that neither children’s nor adult cancer services may fully meet the needs of young people 37 http://bmjopen.bmj.com/ with cancer who characteristically present with a spectrum of rare cancer types requiring 38 39 specialist site specific expertise and additional psychological, educational and social support 40 (8-11). 41 42 Teenage and young adult cancer care is increasingly recognised as an essential specialism. 43 44 In England the National Institute for Health and Clinical Excellence (NICE) published

45 Improving Outcome Guidance for Children and Young People with Cancer in 2005, which on October 1, 2021 by guest. Protected copyright. 46 provided details on how care should be delivered to TYA (12). Central to this guidance was 47 the establishment of specialist TYA Principal Treatment Centres (TYA-PTC) and a mandate 48 that young people aged 15-18 years must receive care in a TYA-PTC and those aged 19-24 49 50 years should have unhindered access to age-appropriate care but could chose to have care 51 more locally to home in an adult cancer unit. Where care is delivered influences clinical 52 outcomes and centralisation of care for rare cancers is advocated. Treatment of cancer in 53 children in a limited number of UK centres since the 1960s contributed to improvements in 54 survival (13, 14). For young people, place of care will influence access to clinical trials, 55 56 treatment protocol (paediatric versus adult regimens) and access to a wider team 57 specialising in TYA care, all of which could influence outcomes. In England, young people 58 have free access to healthcare and can receive all of their care in a TYA-PTC, or all of their 59 care in a children’s or adult cancer unit, or they can receive care between these institutions 60

3 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 25 BMJ Open

1 2 3 having some care delivered in the TYA-PTC with additional components being delivered by 4 children’s/adults services. The decision-making behind referral into a TYA-PTC or a 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 children’s/adult unit is not fully understood and is likely to be influenced by local pathways 7 and by older TYA being offered a choice. 8 9 Despite a lack of evidence, it has been assumed by professionals and young people 10 themselves that age-appropriate care delivered in a specialist environment will positively 11 12 impact outcomes. Age-appropriate services usually include access to a specialist 13 environment, referral to specialist TYA multi-disciplinary team (MDT) meetings in addition to 14 a cancer-site specific MDT, consideration of clinical trial accrual, provision of age- 15 appropriate information, opportunities to preserve fertility and referral to support services 16 17 such as a TYA clinical nurse specialist, psychologist, social worker or youth support co- 18 ordinator. The componentsFor ofpeer this service review that influence outcomes only are not fully described, 19 although increasingly specialist TYA metrics and potential associated outcomes are being 20 proposed (15-17). Associated outcomes are thought to include survival, quality of life, 21 patient-reported outcomes, long-term effects, psychological and social outcomes. 22 23 24 BRIGHTLIGHT was a National Institute for Health Research funded programme of research 25 with an overarching research question: ‘Do Specialist Services for Teenagers and Young 26 Adults with Cancer Add Value?’ (18, 19). Specifically, it aimed to describe: what was age- 27 appropriate care; what were the key components of such a service; what outcomes were 28 29 impacted; and how much did it cost the National Health Service (NHS), young people and 30 their families. Central to this programme was a cohort of young people aged 13-24 years at 31 cancer diagnosis who were recruited within four months of a new cancer diagnosis and 32 followed for three years: the BRIGHTLIGHT Cohort (18). We previously reported that 33 34 survival of the BRIGHTLIGHT cohort was lower than the population diagnosed over the 35 same time period but not recruited to the cohort, which was unrelated to cancer type (18). 36 We surmised that survival differences between the cohort and the non-recruited population 37 may be related to the recruitment window of four months and therefore young people who http://bmjopen.bmj.com/ 38 were sicker had more contact with their treatment team and more opportunities for 39 40 recruitment. The aim of this study was to determine whether there was any evidence of a 41 causal association between the amount of care received in a TYA-PTC on survival outcomes 42 and documentation of clinical processes of care. 43 44 METHODS 45 on October 1, 2021 by guest. Protected copyright. 46 47 Study design 48 This paper reports results from several data sources: the longitudinal cohort study within 49 BRIGHTLIGHT, a mixed methods programme of research, which obtained data from young 50 people through a bespoke survey (18), clinical report data completed by healthcare 51 52 professionals and Demographics Batch Service data from NHS Digital (18). The location of 53 inpatient care for each participant was identified using routinely collected NHS Hospital 54 Episodes Statistics (HES) data. We then developed a bespoke scale to assign each 55 participant a category of care, see Taylor et al. for detail (18). Young people were assigned 56 57 to a category dependent on the proportion of admitted patient care delivered in a TYA-PTC 58 in the first 12 months after diagnosis: all care delivered in a TYA-PTC (ALL-TYA-PTC), no 59 care in a TYA-PTC (NO-TYA-PTC), and some care delivered in a TYA-PTC with additional 60

4 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 25

1 2 3 care in a children’s or adult cancer unit (SOME-TYA-PTC). Sample size calculations were 4 based on the primary outcome measure of the cohort, quality of life (18). 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 7 Participants and setting 8 The BRIGHTLIGHT cohort comprised young people aged 13-24 years, newly diagnosed 9 with cancer (ICD-10 codes C00-C97) in an English hospital and recruited within four-months 10 of diagnosis. Eligibility criteria were as inclusive as possible so no restriction according to 11 12 language or any sensory impairment that affected communication was applied. The only 13 exclusion criteria were: young people receiving a custodial sentence; if the young person 14 was not anticipated to be alive at the first point of data collection (6-months after diagnosis); 15 recurrence of a previous cancer or they were not capable of completing a survey, e.g. 16 17 sedated and in intensive care. The processes for recruitment are reported in detail 18 elsewhere (18, 20,For 21). BRIGHTLIGHT peer wasreview open to recruitment only in 109 NHS hospitals in 19 England, of whom 97 hospitals recruited at least one young person. Young people were 20 recruited between October 2012 and April 2015 (diagnosed between July 2012 and 21 December 2014). They gave written consent (parent consent also obtained from those less 22 23 than 16 years); the study was approved by London-Bloomsbury NHS Research Ethics 24 Committee and the Confidentiality Advisory Group. 25 26 Data collection 27 We assessed documentation of the following clinical processes: 28 29 30  Histological diagnosis 31  Molecular confirmation of diagnosis (where relevant) 32  Cancer stage or prognostic group defined (for leukaemia, total white blood cell count) 33 34  Initial treatment plan 35  Treatment protocol for systemic therapy and/or for radiotherapy 36  Evidence of MDT communication including children’s, TYA or site specific 37 http://bmjopen.bmj.com/ 38  Assessment by supportive care services based on evidence in notes of contact with 39 clinical nurse specialist plus one other member of the MDT (social worker, youth 40 support coordinator, counsellor, psychologist, dietician, physiotherapist, occupational 41 therapist), 42  A record of fertility being discussed 43 44  A record of consideration for inclusion in a clinical trial

45 on October 1, 2021 by guest. Protected copyright. 46 Survival data were obtained from the Demographic Batch Service at NHS Digital reported up 47 until October 2018. 48 49 To describe patient prognosis at recruitment, an existing scoring system was identified that 50 uses anticipated 5-year survival to form groups of patients with expected survival of greater 51 52 than 80%, 50-80% and less than 50% (22). To measure severity of illness we developed a 53 bespoke scale which accounts for the range of cancer types, staging systems, symptom 54 burden, treatment burden, potential late effects and prognosis. This classifies patients as 55 ‘least severe’, ‘intermediate’ and ‘most severe’ based on their cancer-specific information 56 (18). 57 58 59 Analysis 60

5 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 25 BMJ Open

1 2 3 Analysis was based on a predefined analysis plan using STATA Version 15. For each 4 clinical process outcome, the proportion of patients where the item was found to be 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 documented in clinical records was reported by category of care. Proportions were 7 compared across groups using Chi squared tests (including a trend test). 8 9 Survival time for each participant was calculated from date of diagnosis to date of death or 10 censored at the date last known to be alive up to 29th October 2018. Kaplan Meier survival 11 12 curves were plotted for each category of care (NO-TYA-PTC, SOME-TYA-PTC and ALL- 13 TYA-PTC) and estimates of cumulative survival at 1 to 4 years (with 95% confidence 14 intervals) were calculated. The relationship between survival time and TYA category was 15 investigated using a Cox regression model adjusted for confounding factors identified using 16 17 a casual inference approach and based on the conceptual model underpinning the 18 BRIGHTLIGHT SurveyFor (19) inpeer the form ofreview a Directed Acyclic only Graph (DAG) (Supplemental 19 file; DAGitty software www.dagitty.net). Factors adjusted for were age at diagnosis, sex, type 20 of cancer (leukaemia, lymphoma, brain and central nervous system, bone tumours, 21 sarcoma, germ cell, melanoma, carcinomas, other), socioeconomic status (Index of Multiple 22 23 Deprivation rank (IMD) (23)), severity of cancer (least, intermediate, most (18)), ethnicity 24 (white, other), days in hospital over 12 months since diagnosis, treatment received in 12 25 months since diagnosis (Systemic anticancer therapy (SACT) only, radiotherapy (RT) only, 26 Surgery only, Surgery & SACT/SACT & RT & Surgery/RT & SACT/Surgery & RT/Transplant, 27 Other). Geographical region of treatment (North East, North West, Yorkshire, East 28 29 Midlands, West Midlands, London, South East and South West) was included in the model 30 as a random effect (frailty term). The proportional hazard’s assumptions of the Cox 31 regression model were checked. Models were extended to include interaction terms to 32 investigate whether the association between TYA group and survival was different by age at 33 34 diagnosis (using categories of 13 -18 and 19 - 24 years, and age in years) and tumour type 35 (using categories, haematology and oncology). 36

37 http://bmjopen.bmj.com/ 38 RESULTS 39 40 A total of 5953 incident cases were recorded in England, of which 5835 were eligible to 41 participate and 1,126 young people were recruited to the cohort (19.3%). Valid consent was 42 43 available for 1,114. Participation at each wave of data collection has been previously 44 described (18). Participant characteristics are shown in Table 1. In comparison to TYA 45 diagnosed in the same period but not recruited to the cohort, there was under-representation on October 1, 2021 by guest. Protected copyright. 46 of patients with carcinoma, central nervous system cancers and melanoma, and over- 47 48 representation of patients with leukaemia, lymphoma, germ cell tumours and bone tumours 49 (18) . Two diagnostic groups accounted for 50% of the cohort, (lymphoma 31% and germ 50 cell 19%). 51 52 53 54 55 56 57 58 59 60

6 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from

BMJ Open Page 8 of 25

1 2 3 Table 1: BRIGHTLIGHT Cohort characteristics by level of TYA care category at 12 months from diagnosis. Values are frequency (%) 4 unless stated otherwise. 5 6 Characteristic Level of TYA care at 12 months from diagnosis 7 NO-TYA-PTC SOME-TYA-PTC ALL-TYA-PTC 8 N=359 N=415 N=270 9 Age at diagnosis (years) Mean (SD) 21.11 (3.04) 19.44 (3.36) 19.74 (3.23) 10 11 Gender Male 193 (54%) 224 (54%) 156 (58%) 12 FemaleFor peer review166 (46%) only191 (46%) 114 (42%) 13 Ethnicity* N=351 N=408 N=259 14 White 312 (89%) 344 (84%) 221 (85%) 15 Mixed 9 (3%) 9 (2%) 6 (2%)

16 http://bmjopen.bmj.com/ 17 Asian 18 (5%) 36 (9%) 25 (10%) 18 Black 7 (2%) 11 (3%) 2 (1%) 19 Chinese 0 1 (<1%) 2 (1%) 20 Other 5 (1%) 7 (2%) 3 (1%) 21 Socioeconomic status (IMD N=354 N=404 N=263 22 quintile) 1 – most deprived 85 (24%) 100 (25%) 51 (20%) 23 2 67 (19%) 68 (17%) 48 (18%)

24 3 66 (19%) 83 (21%) on October 1, 2021 by guest. Protected copyright. 51 (19%) 25 4 83 (23%) 77 (19%) 49 (19%) 26 5 – least deprived 53 (15%) 76 (19%) 64 (24%) 27 Marital Status N=250 N=262 N=172 28 Married/civil partnership 9 (4%) 8 (3%) 6 (3%) 29 Cohabiting 43 (17%) 27 (10%) 18 (10%) 30 Single/divorced 198 (80%) 227 (87%) 148 (86%) 31 Current status N=277 N=312 N=193 32 33 Working full/part time 126 (45%) 72 (23%) 43 (22%) 34 In education 61 (22%) 112 (36%) 81 (42%) 35 Other work 6 (2%) 5 (2%) 6 (3%) 36 (apprentice/intern/voluntary) 37 Unemployed 10 (4%) 11 (4%) 7 (4%) 38 Long term sick 39 (14%) 51 (16%) 31 (16%) 39 Not seeking work 35 (13%) 61 (19%) 25 (13%) 40 41 42 7 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from

Page 9 of 25 BMJ Open

1 2 3 Type of cancer (1) Leukaemia 27 (8%) 59 (14%) 53 (20%) 4 Lymphoma 138 (38%) 100 (24%) 96 (36%) 5 CNS 12 (3%) 13 (3%) 17 (6%) 6 Bone 10 (3%) 93 (22%) 9 (3%) 7 8 Sarcomas 10 (3%) 30 (7%) 14 (5%) 9 Germ cell 71 (20%) 75 (18%) 46 (17%) 10 Skin 34 (9%) 1 (<1%) 6 (2%) 11 Carcinomas (not skin) 51 (14%) 41 (10%) 27 (10%) 12 MiscellaneousFor peer specified** review5 (1%) only3 (<1%) 1 (<1%) 13 Unspecified Malignant 1 (<1%) 0 1 (<1%) 14 Severity of illness (18) Least 251 (70%) 180 (43%) 131 (49%) 15 Intermediate 67 (19%) 99 (24%) 80 (30%)

16 Most 41 (11%) 136 (33%) http://bmjopen.bmj.com/ 59 (22%) 17 Prognostic score (23) N=354 N=413 N=270 18 <50% 30 (8%) 76 (18%) 61 (23%) 19 50-80% 70 (20%) 166 (40%) 65 (24%) 20 >80% 254 (72%) 171 (41%) 144 (53%) 21 City*** N=359 N=415 N=270 22 23 Birmingham 54 (15%) 75 (18%) 18 (7%) Bristol 65 (18%) 39 (9%) 8 (3%)

24 on October 1, 2021 by guest. Protected copyright. 25 Cambridge 13 (4%) 8 (2%) 2 (1%) 26 Manchester 32 (9%) 44 (11%) 20 (7%) 27 Merseyside 15 (4%) 13 (3%) 11 (4%) 28 East Midlands 19 (5%) 34 (8%) 73 (27%) 29 Leeds 24 (7%) 38 (9%) 39 (14%) 30 Newcastle 15 (4%) 9 (2%) 33 (12%) 31 Oxford 6 (2%) 5 (1%) 8 (3%) 32 London 84 (23%) 116 (28%) 14 (5%) 33 Sheffield 8 (2%) 13 (3%) 13 (5%) 34 Southampton 24 (7%) 21 (5%) 31 (11%) 35 Region*** N=359 N=415 N=270 36 North East 15 (4%) 9 (2%) 33 (12%) 37 North West 47 (13%) 57 (14%) 31 (11%) 38 Yorkshire 32 (9%) 51 (12%) 52 (19%) 39 East Midlands 19 (5%) 34 (8%) 73 (27%) 40 41 42 8 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from

BMJ Open Page 10 of 25

1 2 3 West Midlands 54 (15%) 75 (18%) 18 (7%) 4 London 84 (23%) 116 (28%) 14 (5%) 5 South East 43 (12%) 34 (8%) 41 (15%) 6 South West 65 (18%) 39 (9%) 8 (3%) 7 Treatment received in the SACT only 111 (31%) 114 (27%) 119 (44%) 8 9 first 12 months since Surgery & SACT 55 (15%) 132 (32%) 49 (18%) 10 diagnosis Surgery only 92 (26%) 20 (5%) 23 (9%) 11 SACT & RT 49 (14%) 61 (15%) 30 (11%) 12 Surgery,For RT &peer SACT review12 (3%) only60 (15%) 24 (9%) 13 Surgery & RT 17 (5%) 9 (2%) 16 (6%) 14 Transplant 9 (3%) 12 (3%) 7 (3%) 15 RT only 7 (2%) 5 (1%) 1 (<1%)

16 Other 7 (2%) 2 (<1%) http://bmjopen.bmj.com/ 1 (<1%) 17 Total days in hospital over Median (IQR), [max, min] 13 (4 to 27) 59 (20 to 103) 29 (11 to 73) 18 12 months [1, 213] [2, 228] [1, 286] 19 Given a choice about where N=288 N=356 N=233 20 to receive treatment?$ Yes 121 (42%) 86 (24%) 48 (21%) 21 No (or < 19 years) 167 (58%) 270 (76%) 185 (79%) 22 23 Long term condition prior to N=277 N=311 N=193 cancer? Yes 20 (7%) 34 (11%) 18 (9%)

24 on October 1, 2021 by guest. Protected copyright. 25 No 257 (93%) 277 (89%) 175 (91%) 26 Time to diagnosis: days Median (IQR), [min, max] N=264 N=304 N=188 27 from 1st symptom 62 (29.5 to 169.5) 65.5 (29.5 to 152.5) 63.5 (25.5 to 151.0) 28 [0, 1340] [0, 959] [0, 1217] 29 Time to diagnosis: number Median (IQR), [min, max] N=274 N=311 N=193 30 of GP visits before 1 (0 to 3) 1 (0 to 3) 2 (1 to 3) 31 diagnosis [0, 20] [0, 20] [0, 40] 32 CNS: central nervous system; GP: General Practitioner; IMD: Index of Multiple Deprivation; IQR: interquartile range; RT: radiotherapy; SACT: systemic anticancer therapy 33 * Wave 1 data was used with missing values completed using available PHE data. 34 ** includes 4 ‘unclassified’ – treated in cancer unit but did not have cancer 35 ***where available based on hospital of diagnosis, for 77 cases based on recruiting hospital. Note: Manchester = Christie, Merseyside = Clatterbridge, London = RMH/UCLH 36 $ Those <19 at diagnosis were assumed not to have been given a choice 37 38 39 40 41 42 9 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 11 of 25 BMJ Open

1 2 3 Overall, 359 (34.3%) patients were in the NO-TYA-PTC group, 415 (39.8%) in the SOME- 4 TYA-PTC group and 270 (25.9%) in the ALL-TYA-PTC. Cancer type varied by category of 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 care, lymphoma was the most common in all groups, (38%, 24%, 36%, NO, SOME, ALL 7 respectively). Leukaemia (20%) was the second most common cancer in the ALL-TYA-PTC, 8 bone (22%) in the SOME-TYA-PTC and germ cell (20%) in the NO-TYA-PTC. There was 9 variability in the distribution of prognosis and severity of illness scores between categories of 10 care, the NO-TYA-PTC having the highest proportion of ‘least severe disease’ 70%, 11 12 compared to 43% in the SOME-TYA-PTC and 49% in the ALL-TYA-PTC. The SOME-TYA- 13 PTC had highest proportion of ’most severe’ disease, 33% compared to 11% and 22% in the 14 NO and ALL groups, respectively. The NO-TYA-PTC group were also older. (Table 1) 15 16 17 The number of days in hospital over the 12 months since diagnosis varied between groups. 18 For the NO-TYA-PTCFor group peerthe total number review of days ranged only from 1 to 213 (median 13, 19 interquartile range (IQR), 4-27), for the SOME-TYA-PTC care group, total days ranged from 20 2 to 228 (median 59, IQR 20-103) and for the ALL-TYA-PTC group the total number of days 21 ranged from 1 to 286 (median 29, IQR 11-73). 22 23 24 Processes of care 25 Clinical records were available for 1,078 young people of which 1,009 were assigned to: NO- 26 TYA-PTC (n=333); SOME-TYA-PTC (n=409) and ALL-TYA-PTC (n=267). (HES data were 27 not available for 69 young people so they could not be assigned a category). The 28 29 comparison of processes of care according to category of care are shown in Table 2. There 30 was no evidence of a difference between the three groups for the documentation of: 31 histological diagnosis, cancer stage or prognosis, consideration for entry into a clinical trial 32 and discussion at an MDT. Those receiving NO-TYA-PTC were more likely to have 33 34 documented discussion in a site-specific MDT but had the lowest proportion with 35 documented discussion in a TYA MDT and children’s MDT. There was no significant 36 difference between documentation of an initial treatment plan but there was a trend that this 37 was more likely to have been recorded with more TYA-PTC care. Young people in NO-TYA- http://bmjopen.bmj.com/ 38 PTC had less frequent documentation of a molecular diagnosis (where molecular analysis 39 40 was appropriate), discussions about fertility and assessments by supportive care services 41 defined as contact with a clinical nurse specialist and one other professional such as youth 42 support coordinator, social worker, psychologist (see methods for complete list). 43 44

45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

10 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from

BMJ Open Page 12 of 25

1 2 3 Table 2: Clinical process outcomes 4 Documentation of: NO-TYA-PTC SOME-TYA-PTC ALL-TYA-PTC P-value: 5 2 6 N=333 N=409 N=267 Chi 7 Trend 8 N Yes No N Yes No N Yes No 9 Histological diagnosis 331 307 (93%) 24 (7%) 407 360 (88%) 47 (12%) 265 240 (91%) 25 (9%) 0.14 10 11 0.31 12 Molecular diagnosis 186 For49 (26%) peer137 (74%) 304 review106 (35%) 198 (65%)only200 87 (44%) 113 (56%) 0.002 13 (where relevant)** 0.02 14 Cancer stage or 333 311 (93%) 22 (7%) 409 383 (94%) 26 (6%) 267 253 (95%) 14 (5%) 0.77 15 prognostic group* 0.50

16 http://bmjopen.bmj.com/ 17 Initial treatment plan 330 291 (88%) 39 (12%) 408 370 (91%) 38 (9%) 265 247 (94%) 18 (7%) 0.11 18 0.04 19 Any MDT communication 331 321 (97%) 10 (3%) 408 392 (96%) 16 (4%) 267 259 (97%) 8 (3%) 0.73 20 0.97 21 22 Children’s MDT 329 34 (10%) 295 (90%) 403 81 (20%) 322 (80%) 265 58 (22%) 207 (78%) <0.001 23 <0.001

24 TYA MDT 326 164 (50%) 162 (50% 401 285 (71%) 116 (29%) 265 207 on October 1, 2021 by guest. Protected copyright. (78%) 58 (22%) <0.001 25 <0.001 26 27 Site specific MDT 325 271 (83%) 54 (17%) 402 285 (71%) 117 (29%) 264 189 (72%) 75 (28%) <0.001 28 0.001 29 Assessment by 327 124 (38%) 203 (62%) 405 249 (61%) 156 (39%) 258 154 (60%) 104 (40%) <0.001 30 supportive care services <0.001 31 32 Fertility being discussed 330 178 (54%) 152 (46%) 407 282 (69%) 125 (31%) 259 195 (75%) 64 (25%) <0.001 33 (all) <0.001 34 Fertility discussed 178 112 (63%) 66 (27%) 221 172 (78%) 49 (22%) 152 117 (77%) 35 (23%) 0.002 35 (Males) 0.003 36 37 Fertility discussed 152 66 (43%) 110 (59%) 186 110 (59%) 76 (41%) 107 78 (73%) 29 (27%) <0.001 38 (Females) <0.001 39 40 41 42 11 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from

Page 13 of 25 BMJ Open

1 2 3 Consideration into 328 207 (63%) 121 (37%) 405 252 (62%) 153 (38%) 256 176 (69%) 80 (31%) 0.21 4 clinical trial 0.19 5 6 From case report form (CRF) data: completed/partially completed=1078; 1009 have category of specialist care recorded TYA: teenage and young adult; MDT: multidisciplinary team 7 * Cancer stage or prognostic group documented is defined as: For leukaemia – a WBC (white blood cell count) measure is provided; For lymphoma if stage (1-4) is entered (variable "stage"); For solid 8 tumour use variable "has the tumour been staged?" If these things are not recorded for the appropriate cancer type, then coded as not documented. Cancer type is determined by birch classification. 9 ** indicated as “not relevant” for: NO-TYA-PTC, n=137; SOME-TYA-PTC, n=97; ALL-TYA-PTC, n= 65 10 11 12 For peer review only 13 14 15

16 http://bmjopen.bmj.com/ 17 18 19 20 21 22 23

24 on October 1, 2021 by guest. Protected copyright. 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 12 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 14 of 25

1 2 3 Survival 4 The duration of follow-up by October 2018 is shown in Table 3. The number of deaths in the 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 NO-TYA-PTC group was 27 (8%), compared to 35 (13%) in ALL-TYA-PTC and 91 (22%) in 7 SOME-TYA-PTC. The cumulative probability of survival by time since diagnosis for the TYA- 8 PTC categories is shown in Figure 1 and Table 4. Although survival probabilities at one year 9 were similar there was clear divergence between the groups over the following time period, 10 such that probabilities were highest for those receiving NO-TYA-PTC, followed by ALL-TYA- 11 12 PTC care, then SOME-TYA-PTC care. Following full adjustment for confounding factors, 13 regression (Table 5) showed there was no evidence of a relationship between the category of 14 care and hazard (risk) of death. 15 16 17 Table 3: Duration of follow-up 18 ForNO-TYA-PTC peerSOME-TYA-PTC review onlyALL-TYA-PTC ALL 19 N=359 N=415 N=270 N=114 20 Median (IQR) 1839 1743 1747 1779 21 follow up (days) (1597 to 2041) (1474 to 1991) (1536 to 2023) (1536 to 2023) 22 23 24 25 Table 4: Estimated cumulative survival probabilities by categories of TYA care and year 26 from diagnosis (95% CI) 27 28 NO-TYA-PTC SOME-TYA-PTC ALL-TYA-PTC 29 1 year 0.98 (0.96 to 0.99) 0.97 (0.95 to 0.99) 0.98 (0.95 to 0.99) 30 2 years 0.95 (0.92 to 0.97) 0.89 (0.86 to 0.92) 0.93 (0.89 to 0.95) 31 3 years 0.94 (0.91 to 0.96) 0.83 (0.79 to 0.86) 0.90 (0.85 to 0.93) 32 33 4 years 0.93 (0.90 to 0.95) 0.80 (0.76 to 0.84) 0.89 (0.84 to 0.92) 34 35 36

37 Table 5: Results from Cox regression model for survival from diagnosis by categories http://bmjopen.bmj.com/ 38 of TYA care received during the first 12 months from diagnosis 39 Hazard 95% Confidence P-value * 40 Ratio Interval 41 Unadjusted model (N=1044) 42 43 TYA care category SOME-TYA-PTC 3.14 2.04 to 4.83 P<0.001 44 (v NO-TYA-PTC) ALL-TYA-PTC 1.79 1.08 to 2.96 45 Fully adjusted model (N=1000)** on October 1, 2021 by guest. Protected copyright. 46 TYA care category SOME-TYA-PTC 1.55 0.94 to 2.58 P=0.15 47 48 (v NO-TYA-PTC) ALL-TYA-PTC 1.13 0.64 to 1.97 49 * P-value from a likelihood ratio test 50 ** adjusted for age at diagnosis, sex, type of cancer, socioeconomic status (IMD rank), severity of cancer, treatment, days in 51 hospital, and ethnicity Geographical region of treatment was included as a random effect (frailty term). 52 53 Subgroup analyses showed no statistical evidence that the relationship between survival and 54 level of care was different for the combined group of leukaemia and lymphomas compared 55 with other cancers (Table 6). There was however weak evidence of a difference in the effect 56 of level of care on survival by age group, notably with lower risk of death when comparing 57 58 SOME-TYA-PTC and ALL-TYA-PTC with NO-TYA-PTC in those aged under 19 years at 59 diagnosis, while these relative risks were higher in the over 19 group. A similar pattern was 60 seen when considering age as continuous.

13 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 25 BMJ Open

1 2 3 4 Table 6: Planned subgroup investigations for cancer type (leukaemia/lymphoma v 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from other) and age group (<19 v 19+): Results from fully adjusted* models with interaction 6 7 terms (N=1000) 8 TYA care Fully adjusted 95% confidence P-value from 9 category hazard ratio interval interaction 10 Cancer type 11 Leukaemia/ SOME-TYA-PTC v 1.37 0.63 to 3.01 P=0.95 12 13 Lymphoma NO-TYA-PTC 14 ALL-TYA-PTC v 0.97 0.41 to 2.28 15 NO-TYA-PTC 16 Other cancers SOME-TYA-PTC v 1.34 0.68 to 2.63 17 NO-TYA-PTC 18 For peer review only 19 ALL-TYA-PTC v 1.09 0.52 to 2.27 20 NO-TYA-PTC 21 Age group 22 23 Age <19 years SOME-TYA-PTC v 0.81 0.41 to 1.57 P=0.08 24 NO-TYA-PTC 25 ALL-TYA-PTC v 0.79 0.37 to 1.71 26 NO-TYA-PTC 27 Age 19+ years SOME-TYA-PTC v 1.75 0.99 to 3.06 28 29 NO-TYA-PTC 30 ALL-TYA-PTC v 1.14 0.59 to 2.23 31 NO-TYA-PTC 32 Continuous Coefficient for 33 34 age age (per year) 35 NO-TYA-PTC 0.95 0.85 to 1.06 P=0.07 36 SOME-TYA-PTC 1.11 1.03 to 1.18 http://bmjopen.bmj.com/ 37 ALL-TYA-PTC 1.05 0.94 to 1.17 38 39 * adjusted for age at diagnosis, type of cancer (detailed categories), socioeconomic status (IMD rank), severity of cancer, ethnicity (white v other), gender, treatment (detailed categories) received in 6 months from diagnosis, 40 days in hospital within 12 months of diagnosis with region as random effect. 41 42 43 44 DISCUSSION 45 on October 1, 2021 by guest. Protected copyright. 46 We have reported on a national longitudinal evaluation of specialist cancer services for 47 young people aged 13-24 years at diagnosis defining the TYA-PTCs and their networks as 48 they were described in the UK NICE Improving Outcomes Guidance in 2005 (12). We used 49 routinely collected NHS data (HES) which records hospital admission data to measure how 50 much care young people received in the TYA-PTC, dividing our cohort into three distinct 51 52 groups, all care delivered in a TYA-PTC (ALL-TYA-PTC) no care in a TYA-PTC (NO-TYA- 53 PTC) and those who received some care in the TYA-PTC and other parts of their care in 54 another children’s or adult hospital. We assessed documentation of clinical processes 55 assumed to be related to quality of care and found those receiving NO-TYA-PTC were less 56 likely to have a record of molecular diagnosis (where relevant). Additionally, this group were 57 less likely to have documentation of review by a children’s or TYA MDT, have an 58 assessment by supportive care services or have a fertility discussion compared to those 59 60 treated in SOME-TYA-PTC or ALL-TYA-PTC. These are criteria which we would expect to

14 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 25

1 2 3 be associated with specialist age-appropriate care (16,17), therefore it is not surprising these 4 appear to be more frequently documented in the ALL and SOME group. 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 Our results suggest differences between the groups in these measures of the quality of 7 cancer care delivered to young people. However, this did not appear to have an impact on 8 survival outcomes. One-year survival was similar between the groups; at four years, survival 9 10 was highest in the NO-TYA-PTC group, followed by ALL-TYA-PTC and lowest in the SOME- 11 TYA-PTC group. However, this was not significant. One of the reasons for implementing a 12 new model of care in 2005 specific for TYA with cancer was due to the disparity in survival 13 compared to children and older adults (12). It is therefore disappointing that there were no 14 survival differences noted between the three categories. Alternatively, we could view this as 15 a positive finding - wherever young people choose to be treated, their survival outcomes are 16 the same. The work we did with young people to develop BRIGHTLIGHT highlighted that 17 they did not perceive survival alone as the most import outcome. Quality of life and the ability 18 For peer review only 19 to get back on with life were as important (18,19). While we have shown a better 20 improvement in quality of life when treated in a TYA-PTC (Taylor et al. under review), we 21 have yet to ascertain whether young people's reintegration into life when treatment ends is 22 also better. 23 24 There was weak evidence that increasing age was associated with higher risk of death for 25 those in the SOME-TYA-PTC compared to NO-TYA-PTC and ALL-TYA-PTC, an important 26 finding given that it is this group who have a choice over where to receive their care (12). We 27 believe that further investigation into the lower survival in the SOME-TYA-PTC group is 28 warranted, particularly the association with age. Those aged over 19 years at diagnosis had 29 30 the more pronounced effects in the subgroup analysis, but the direction of the effect differed 31 substantially for under 19s compared to the overall effect. 32 33 There is a paucity of existing literature to compare our results with and comparisons are 34 further confounded by variation in healthcare systems, distinct models of specialist age- 35 appropriate care adopted and the international definition of TYA which can extend up to 39 36 years in some countries (24). A previous retrospective regional study of children and TYA in 37 England found a survival benefit of being treated in a PTC for poor prognosis leukaemia and http://bmjopen.bmj.com/ 38 a converse relationship for those with soft tissue sarcoma, no significant differences in 39 survival were observed for those with lymphoma, CNS, bone and germ cell tumours (25). Of 40 note, a previous study has also shown those receiving ‘SOME’ specialist care have poorer 41 42 survival for some indications (26). These studies suggest that some tumour groups may 43 benefit from care at the PTC however, due to our previously reported difficulties with 44 recruitment and reduction in sample size (20) we were unable to conduct the detailed 45 analysis of individual cancer types as planned, and thus benefits of the PTC may be masked on October 1, 2021 by guest. Protected copyright. 46 within the grouping of ‘haematology’ and ‘solid tumours’. 47 48 Limitations 49 50 Despite our study including a large, broadly representative sample of newly diagnosed TYA 51 with cancer followed-up for 3 years, and our analyses being adjusted for factors known to 52 affect outcomes in cancer there are some limitations to our study. Our definition of 53 ‘specialist’ was based on the TYA-PTC care model as defined by the NICE Improving 54 Outcomes Guidance Issued in 2005 (12), which does not necessarily reflect current delivery 55 of age-appropriate care (9). The study population were recruited during a period of evolution 56 57 of TYA services in England, therefore the models of care are unlikely to reflect current 58 practice, particularly as we have identified that specialist age-appropriate care takes time to 59 develop (13). Additionally, categorising TYA-PTC assumes that all PTCs are equal and does 60 not measure the quality of care delivered. We know national variation exists in configuration

15 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 25 BMJ Open

1 2 3 and maturity of services (27), particularly during 2012-2014 when patients were recruited. 4 Further, due to the coding of hospital inpatient data it is possible that some patients have 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from been misattributed as receiving care in the TYA-PTC when they may have been cared for in 6 7 a Trust which had a TYA-PTC but care was delivered at a different hospital and not in the 8 TYA unit. 9 10 Consideration must also be given to additional factors influencing survival outcomes which 11 we did not collect or measure. These include deviation from the intended treatment plan 12 such as the proportion of treatment delivered, delays/reduction in delivery and toxicity. 13 Therefore, it was not possible to determine the dose or type of chemotherapy or 14 radiotherapy received by patients in each group and these would be important determinants 15 of survival. Our scale for place of care was derived only from inpatient care and not care or 16 treatment delivered as an outpatient. Thus, we may have missed considerable elements of 17 the care received. 18 For peer review only 19 Overall survival of the cohort was lower than those diagnosed during the same period but 20 not recruited and so may not reflect the experience of the whole population (18), we do not 21 22 know the decision making processes behind referral of patients into each TYA-PTC group at 23 diagnosis, it may be that patients with better prognosis are treated more locally with site 24 specific expertise competent at treating the cancer with good survival outcomes and those 25 with more complex disease and holistic needs are referred into the specialist TYA service. 26 27 Conclusion 28 29 We have reported on the first systematic longitudinal evaluation of cancer services for young 30 people. Overall, survival at four years was good across all three categories of care with 31 some differences between the NONE, ALL and SOME groups as defined by NICE improving 32 outcomes service specification in 2005. The factors contributing to survival differences 33 between the groups warrants further investigation particularly the relationship between 34 survival, level of TYA care and age. BRIGHTLIGHT results are immediately important for 35 current healthcare recommendations for young people with cancer in England. The currently 36

37 proposed model of care advocates ‘Joint Care’ between the TYA-PTCs and selected local http://bmjopen.bmj.com/ 38 hospitals, further enquiry is required with additional data collection to assess whether this 39 ‘Joint Care’ would generate a similar pattern of survival trends as the ‘SOME-TYA-PTC’ 40 Group in our study. 41 42 43 44 REFERENCES on October 1, 2021 by guest. Protected copyright. 45 1. Birch JM, Alston RD, Kelsey AM, Quinn MJ, Babb P, McNally RJQ. Classification and 46 incidence of cancers in adolescents and young adults in England 1979–1997. British 47 48 Journal of Cancer. 2002;87(11):1267-74. 49 2. Fern LA, Birch R, Whelan J, Cooke M, Sutton S, Neal RD, et al. Why can't we 50 improve the timeliness of cancer diagnosis in children, teenagers, and young adults? 51 BMJ. 2013;347:f6493. 52 3. Fern LA, Lewandowski JA, Coxon KM, Whelan J, National Cancer Research Institute 53 T, Young Adult Clinical Studies Group UK. Available, accessible, aware, appropriate, 54 and acceptable: a strategy to improve participation of teenagers and young adults in 55 cancer trials. Lancet Oncol. 2014;15(8):e341-50. 56 4. Barr RD, Ries LA, Lewis DR, Harlan LC, Keegan TH, Pollock BH, et al. Incidence 57 and incidence trends of the most frequent cancers in adolescent and young adult 58 Americans, including "nonmalignant/noninvasive" tumors. Cancer. 2016;122(7):1000- 59 8. 60

16 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 25

1 2 3 5. Tricoli JV, Blair DG, Anders CK, Bleyer WA, Boardman LA, Khan J, et al. Biologic 4 and clinical characteristics of adolescent and young adult cancers: Acute 5 lymphoblastic leukemia, colorectal cancer, breast cancer, melanoma, and sarcoma. BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 Cancer. 2016;122(7):1017-28. 7 6. Herbert A, Lyratzopoulos G, Whelan J, Taylor RM, Barber J, Gibson F, et al. 8 Diagnostic timeliness in adolescents and young adults with cancer: a cross-sectional 9 10 analysis of the BRIGHTLIGHT cohort. Lancet Child Adolesc Health. 2018;2(3):180- 11 90. 12 7. Boissel N, Baruchel A. Acute lymphoblastic leukemia in adolescent and young adults: 13 treat as adults or as children? Blood. 2018;132(4):351-61. 14 8. Fern LA, Taylor RM, Whelan J, Pearce S, Grew T, Brooman K, et al. The art of age- 15 appropriate care: reflecting on a conceptual model of the cancer experience for 16 teenagers and young adults. Cancer Nurs. 2013;36(5):E27-38. 17 9. Lea S, Taylor RM, Martins A, Fern LA, Whelan JS, Gibson F. Conceptualizing age- 18 appropriate Forcare for teenagerspeer and review young adults with only cancer: a qualitative mixed- 19 methods study. Adolesc Health Med Ther. 2018;9:149-66. 20 10. Barr RD. Planning a Comprehensive Program in Adolescent and Young Adult 21 Oncology-A Collision with Reality. J Adolesc Young Adult Oncol. 2016;5(4):303-9. 22 11. Morgan S, Davies S, Palmer S, Plaster M. Sex, Drugs, and Rock ‘n’ Roll: Caring for 23 Adolescents and Young Adults With Cancer. Journal of Clinical Oncology. 24 2010;28(32):4825-30. 25 12. National Institute for Health and Care Excellence. Guidance on cancer services: 26 27 improving outcomes in children and young people with cancer. NICE, London 2005. 28 https://www.nice.org.uk/guidance/csg7/resources/improving-outcomes-in-children- 29 and-young-people-with-cancer-update-773378893 [Accessed 26/08/2020] 30 13. Stiller CA. Centralisation of treatment and survival rates for cancer. Arch Dis Child. 31 1988;63(1):23-30. 32 14. Stiller CA. Centralised treatment, entry to trials and survival. Br J Cancer. 33 1994;70(2):352-62. 34 15. Greenberg M, Klassen A, Gafni A, McBride ML, Albritton K. Outcomes and metrics: 35 measuring the impact of a comprehensive adolescent and young adult cancer 36

37 program. Cancer. 2011;117(10 Suppl):2342-50. http://bmjopen.bmj.com/ 38 16. Rae CS, Pole JD, Gupta S, Digout C, Szwajcer D, Flanders A, et al. Development of 39 System Performance Indicators for Adolescent and Young Adult Cancer Care and 40 Control in Canada. Value Health. 2020;23(1):74-88. 41 17. Sironi G, Barr RD, Ferrari A. Models of Care-There Is More Than One Way to 42 Deliver. Cancer J. 2018;24(6):315-20. 43 18. Taylor RM, Fern LA, Barber J, Alvarez-Galvez J, Feltbower R, Morris S, et al. 44 Description of the BRIGHTLIGHT cohort: the evaluation of teenage and young adult

45 cancer services in England. BMJ Open. 2019;9(4):e027797. on October 1, 2021 by guest. Protected copyright. 46 19. Taylor RM, Fern LA, Solanki A, Hooker L, Carluccio A, Pye J, et al. Development and 47 validation of the BRIGHTLIGHT Survey, a patient-reported experience measure for 48 young people with cancer. Health Qual Life Outcomes. 2015;13:107. 49 20. Kenten C, Martins A, Fern LA, Gibson F, Lea S, Ngwenya N, et al. Qualitative study 50 to understand the barriers to recruiting young people with cancer to BRIGHTLIGHT: 51 a national cohort study in England. BMJ Open. 2017;7(11):e018291. 52 21. Taylor RM, Fern LA, Aslam N, Whelan JS. Direct access to potential research 53 participants for a cohort study using a confidentiality waiver included in UK National 54 Health Service legal statutes. BMJ Open. 2016;6(8):e011847. 55 22. Husson O, Zebrack BJ, Block R, Embry L, Aguilar C, Hayes-Lattin B, et al. Health- 56 57 Related Quality of Life in Adolescent and Young Adult Patients With Cancer: A 58 Longitudinal Study. J Clin Oncol. 2017;35(6):652-9. 59 60

17 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 25 BMJ Open

1 2 3 23. Department for Communities and Local Government. The English Indices of 4 Deprivation 2015 Statistical Release. England: Department for Communities and 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 Local Government 2015. 7 24. Pini SA, Gibson F, Fern LA, Morgan SJ, Phillips RS, Stark DP. Multi-Professional 8 Perspectives on Adolescent and Young Adult Oncology Across Europe: An e-Delphi 9 Survey. J Adolesc Young Adult Oncol. 2017;6(1):178-85. 10 25. Fairley L, Stark DP, Yeomanson D, Kinsey SE, Glaser AW, Picton SV, et al. Access 11 to principal treatment centres and survival rates for children and young people with 12 cancer in Yorkshire, UK. BMC Cancer. 2017;17(1):168. 13 26. Birch R. Teeange and Young Adult Cancer in England - The patient journey and 14 experience 15 http://etheses.whiterose.ac.uk/4157/1/Birch_RJ_Faculty_of_Medicine_and_Health_P 16 hD_2013.pdf: University of Leeds; 2013. 17 27. Vindrola-Padros C, Taylor RM, Lea S, Hooker L, Pearce S, Whelan J, et al. Mapping 18 Adolescent ForCancer Services:peer How review Do Young People, only Their Families, and Staff 19 Describe Specialized Cancer Care in England? Cancer Nurs. 2016;39(5):358-66. 20 21 22 Author contributions: 23 RMT, LAF, JB, DS, SM, RF, LH, FG, RR, JSW were involved in developing the protocol. 24 RMT coordinated the running of the study and was responsible for data acquisition. JB, JGA, 25 RMT, LAF, SM, RF, DS, JSW contributed to the analysis. RMT, LAF, JB and JSW drafted 26 27 the manuscript. All authors critically revised and approved the final manuscript. 28 29 Patient and public involvement: 30 Young people have been involved in this study from the feasibility stage onward. They were 31 involved in study development, acted as co-researchers and were instrumental in the design 32 33 and methods of the study. A representative of the Young Advisory Panel (YAP) was a co- 34 applicant on the grant and the YAP have been part of the management of the study since 35 the grant was awarded in 2011. Details of the extent of young people’s involvement in 36 BRIGHTLIGHT is provided in reference 13. 37 http://bmjopen.bmj.com/ 38 39 Acknowledgements: 40 We would like to thank the members of our Young Advisory Panel (Zeena Beale, Ciaran 41 Fenton, Emily Freemantle, Jaasjan Guvindia, Laura Haddard, Steph Hammersley, Amy 42 Lang, Joshua Lerner, Tanya Loughlin, Jason, Sin Jin Loo, Jennifer Miller, Maria Onasanya, 43 44 Arif Nasir, Steph Still, Poppy Richards, Amy Riley, Lara Veitch, Freya Voss, JJ Wheeler,

45 Max Willliamson, Antonia Young), the 1,114 young people who consented to participate in on October 1, 2021 by guest. Protected copyright. 46 BRIGHTLIGHT, healthcare professionals who approached and consented young people, 47 and ex-members of the team who have contributed to study management (Catherine 48 O’Hara, Anita Solanki, Natasha Aslam, Zuwena Fox). 49 50 51 We would like to dedicate this manuscript in memory of Mr Stephen Sutton and Mr Mathew 52 Cook who were instrumental to study set up, design and management. Both of whom died 53 from their cancer during the study. 54 55 56 We would also like to thanks the following for all their support with recruitment to 57 BRIGHTLIGHT: the National Cancer Research Institute, especially Dr Eileen Loucaides and 58 the Secretariat; Matt Seymour, Matt Cooper and Karen Poole at the former National Cancer 59 Research Network; Maria Khan and Sabrina Sandhu (North West Knowledge Intelligence 60

18 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 25

1 2 3 Team); TYAC; Teenage Cancer Trust; CLIC Sargent; Ipsos MORI; Quality Health and the 4 research teams at 109 NHS Trusts in England who opened BRIGHTLIGHT to recruitment. 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 7 Principal Investigators agreeing to be acknowledged for their contribution to BRIGHTLIGHT 8 recruitment: 9 Claire Hemmaway, Barking, Havering and Redbridge Hospitals NHS Trust; Anita Amadi, 10 Barnet and Chase Farm Hospitals NHS Trust; Keith Elliott, Barnsley Hospital NHS 11 12 Foundation Trust; Leanne Smith, Blackpool, Fylde and Wyre Hospitals NHS Trust; Shirley 13 Cocks, Bolton NHS Foundation Trust; Victoria Drew, Bradford Teaching Hospitals NHS 14 Foundation Trust; Elizabeth Pask, Central Manchester University Hospitals NHS Foundation 15 Trust; Anne Littley, Central Manchester University Hospitals NHS Foundation Trust; Mark 16 17 Bower, Chelsea and Westminster Hospital NHS Trust; Scott Marshall, City Hospitals 18 Sunderland NHS FoundationFor peer Trust; Lorna review Dewar, Colchester only Hospital University NHS Trust; 19 Nnenna Osuji, Croydon Health Services NHS Trust; David Allotey, Derby Hospitals NHS 20 Foundation Trust; Karen Jewers, East Lancashire Hospitals NHS Trust; Asha Johny, 21 Gloucestershire Hospitals NHS Foundation Trust; Nicola Knightly, Great Western Hospitals 22 23 NHS Foundation Trust; Robert Carr, Guy's & St Thomas' Hospital NHS Foundation Trust; 24 Alison Milne, Hampshire Hospitals NHS Foundation Trust; Claire Hall, Harrogate and District 25 NHS Foundation Trust; James Bailey, Hull and East Yorkshire Hospitals NHS Trust; 26 Christine Garlick, Ipswich Hospital NHS Foundation Trust; Alison Brown, Isle of Wight 27 Healthcare NHS Trust; Carolyn Hatch, Lancashire Teaching Hospitals NHS Foundation 28 29 Trust; Vivienne E. Andrews, Medway NHS Foundation Trust; Sara Greig, Milton Keynes 30 Hospital NHS Foundation Trust; Jennifer Wimperis, Norfolk and Norwich University Hospital 31 NHS Trust; Suriya Kirkpatrick, North Bristol NHS Trust; Jonathan Nicoll, North Cumbria 32 University Hospitals NHS Trust; Ivo Hennig, Nottingham University Hospitals NHS Trust; 33 34 Karen Sherbourne, Oxford Radcliffe Hospital NHS Trust; Clare Turner, Plymouth Hospitals 35 NHS Trust; Claire Palles-Clark, Royal Surrey County Hospital NHS Trust; Christine Cox, 36 Royal United Hospital Bath NHS Trust; Yeng Ang, Salford Royal NHS Foundation Trust; 37 Jonathan Cullis, Salisbury NHS Foundation Trust; Daniel Yeomanson, Sheffield Children's http://bmjopen.bmj.com/ 38 NHS Foundation Trust; Ruth Logan, Sheffield Teaching Hospitals NHS Foundation Trust; 39 40 Deborah Turner, South Devon Healthcare NHS Trust; Dianne Plews, South Tees Hospitals 41 NHS Trust; Juliah Jonasi, Southend University Hospital NHS Foundation Trust; Ruth 42 Pettengell, St George's Healthcare NHS Trust; Kamal Khoobarry, Surrey & Sussex 43 Healthcare NHS Trust; Angela Watts, The Dudley Group of Hospitals NHS Foundation Trust; 44 Louise Soanes, The Royal Marsden NHS Foundation Trust; Claudette Jones, The Royal 45 on October 1, 2021 by guest. Protected copyright. 46 Orthopaedic Hospital NHS Trust; Michael Jenkinson, The Walton Centre for Neurology and 47 Neurosurgery NHS Trust; Nicky Pettitt, University Hospital Birmingham NHS Foundation 48 Trust; Vijay Agarwal, University Hospital Birmingham NHS Foundation Trust; Beth Harrison, 49 University Hospitals Coventry and Warwickshire NHS Trust; Fiona Miall, University Hospitals 50 of Leicester NHS Trust; Gail Wiley, University Hospitals of Morecambe Bay NHS Trust; 51 52 Lynda Wagstaff, Walsall Hospitals NHS Trust; Fiona Smith, West Hertfordshire Hospitals 53 NHS Trust; Sarah Janes, Western Sussex NHS Trust; Serena Hillman, Weston Area Health 54 NHS Trust; Christopher Zaborowski, Yeovil District Hospital NHS Foundation Trust. 55 56 57 Data for this study is based on information collected and quality assured by the PHE 58 National Cancer Registration and Analysis Service. Access to the data was facilitated by the 59 PHE Office for Data Release. 60

19 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 25 BMJ Open

1 2 3 Conflict of interest: 4 None declared. 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 7 Ethics Approval: 8 The study was approved by the Health Research Authority Confidentiality Advisory Group 9 (reference ECC 8-05(d)/2011) and London Bloomsbury NHS Research Ethics Committee 10 (reference LO/11/1718). 11 12 13 Funding: 14 This paper presents independent research funded by the National Institute for Health 15 Research (NIHR) under its Programme Grants for Applied Research Programme (Grant 16 17 Reference Number RP-PG-1209-10013). The views expressed are those of the author(s) 18 and not necessarilyFor those of peerthe NIHR or review the Department ofonly Health and Social Care. The 19 BRIGHTLIGHT Team acknowledges the support of the NIHR, through the Cancer Research 20 Network. 21 22 23 LAF and LH are funded by Teenage Cancer Trust, DS holds research grant funding from 24 Teenage Cancer Trust, and RR was (in part) supported by the National Institute for Health 25 Research (NIHR) Collaboration for Leadership in Applied Health Research and Care 26 (CLAHRC) North Thames at Bart’s Health NHS Trust. RMT is a National Institute for Health 27 Research (NIHR) Senior Nurse Research Leader. The views expressed in this article are 28 29 those of the author and not necessarily those of the NIHR, or the Department of Health and 30 Social Care. JAG was subsidised by the Ramon & Cajal programme operated by the 31 Ministry of Economy and Business (RYC-2016-19353), and the European Social Fund. None 32 of the funding bodies have been involved with study concept, design or decision to submit 33 34 the manuscript. 35 36 Data sharing statement: 37 Further details of the BRIGHTLIGHT programme of work are available through the study http://bmjopen.bmj.com/ 38 website (www.brightlightstudy.com). Data that are not held under licence with Public Health 39 40 England or NHS Digital will be available when the primary analysis is complete. We welcome 41 collaboration, for general data sharing enquiries please contact RMT ([email protected]). 42 43 44

45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

20 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 25

1 2 3 Figure legends 4 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 Figure 1: Comparison of survival according to the three categories of care (unadjusted) 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44

45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

21 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 25 BMJ Open

1 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 151x104mm (150 x 150 DPI) 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on October 1, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from

BMJ Open Page 24 of 25

1 2 3 Supplemental file 4 5 6 Figure A1: Matrix representing the causal diagram explaining confounding variable to enter in the Directed Acyclic Graph (DAG) 7 TYA QOL Survival Age Gender CS CT SES Ethnicity Geography DoH Treatment RtD 8 CoC 9 TYA CoC         O    10 Survival            11 12 Age ForO peer  review O O only    13 Gender   O O O     14 CS         15 CT       

16 http://bmjopen.bmj.com/ 17 SES       18 Ethnicity      19 Geography     20 DoH 21    22 Treatment   23 RtD 

24 QOL on October 1, 2021 by guest. Protected copyright. 25 26 CS: cancer severity; CT: cancer type; DoH: duration of hospitalisation; QOL: quality of life; RtD: route to diagnosis; SE: symptom experience; SES: 27 socioeconomic status; SS: social support; TYA CoC: teenage and young adult category of care 28 O – indicates a null relationship 29 30 31 32 33 34 35 36 37 38 39 40 41 42 1 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from

Page 25 of 25 BMJ Open

1 2 STROBE 2007 (v4) Statement—Checklist of items that should be included in reports of cohort studies 3 4 5 Item Section/Topic Recommendation Reported on page # 6 # 7 Title and abstract 1 (a) Indicate the study’s design with a commonly used term in the title or the abstract 1 8 9 (b) Provide in the abstract an informative and balanced summary of what was done and what was found 2-3 10 Introduction 11 12 Background/rationale 2 Explain theFor scientific background peer and rationale reviewfor the investigation being reported only 3-4 13 Objectives 3 State specific objectives, including any prespecified hypotheses 4 14 15 Methods

16 Study design 4 Present key elements of study design early in the paper http://bmjopen.bmj.com/ 4-5 17 18 Setting 5 Describe the setting, locations, and relevant dates, including periods of recruitment, exposure, follow-up, and data 5 19 collection 20 Participants 6 (a) Give the eligibility criteria, and the sources and methods of selection of participants. Describe methods of follow-up 5 (reference cited for 21 details) 22 23 (b) For matched studies, give matching criteria and number of exposed and unexposed na Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, and effect modifiers. Give diagnostic criteria, if 5 24 on October 1, 2021 by guest. Protected copyright. 25 applicable 26 Data sources/ 8* For each variable of interest, give sources of data and details of methods of assessment (measurement). Describe 5 27 28 measurement comparability of assessment methods if there is more than one group 29 Bias 9 Describe any efforts to address potential sources of bias 6 30 Study size 10 Explain how the study size was arrived at 5 31 32 Quantitative variables 11 Explain how quantitative variables were handled in the analyses. If applicable, describe which groupings were chosen and 6 33 why 34 Statistical methods 12 (a) Describe all statistical methods, including those used to control for confounding 6 35 36 (b) Describe any methods used to examine subgroups and interactions 6 37 (c) Explain how missing data were addressed 6 38 (d) If applicable, explain how loss to follow-up was addressed 6 39 40 (e) Describe any sensitivity analyses 6 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from

BMJ Open Page 26 of 25

1 2 Results 3 4 Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers potentially eligible, examined for eligibility, confirmed 6, reference cited 5 eligible, included in the study, completing follow-up, and analysed 6 (b) Give reasons for non-participation at each stage Referenced cited 7 8 (c) Consider use of a flow diagram Referenced cited 9 Descriptive data 14* (a) Give characteristics of study participants (eg demographic, clinical, social) and information on exposures and potential 6-9 10 confounders 11 (b) Indicate number of participants with missing data for each variable of interest Referenced cited 12 For peer review only 13 (c) Summarise follow-up time (eg, average and total amount) Referenced cited 14 Outcome data 15* Report numbers of outcome events or summary measures over time Referenced cited 15 Main results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted estimates and their precision (eg, 95% confidence 10-12

16 http://bmjopen.bmj.com/ interval). Make clear which confounders were adjusted for and why they were included 17 18 (b) Report category boundaries when continuous variables were categorized na 19 (c) If relevant, consider translating estimates of relative risk into absolute risk for a meaningful time period na 20 Other analyses 17 Report other analyses done—eg analyses of subgroups and interactions, and sensitivity analyses 12-14 21 22 Discussion 23 Key results 18 Summarise key results with reference to study objectives 14

24 on October 1, 2021 by guest. Protected copyright. 25 Limitations 26 Interpretation 20 Give a cautious overall interpretation of results considering objectives, limitations, multiplicity of analyses, results from 15-16 27 similar studies, and other relevant evidence 28 Generalisability 21 Discuss the generalisability (external validity) of the study results 15-16 29 30 Other information 31 Funding 22 Give the source of funding and the role of the funders for the present study and, if applicable, for the original study on 20 32 which the present article is based 33 34 35 *Give information separately for cases and controls in case-control studies and, if applicable, for exposed and unexposed groups in cohort and cross-sectional studies. 36 37 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and published examples of transparent reporting. The STROBE 38 39 checklist is best used in conjunction with this article (freely available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at 40 http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is available at www.strobe-statement.org. 41 42 43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from

Article Type: Original research

Date Submitted by the 08-Feb-2021 Author:

Whelan, Jeremy; University College London Hospitals NHS Foundation on October 1, 2021 by guest. Protected copyright. Trust,

Primary Subject Oncology Heading:

Secondary Subject Heading: Communication, Health services research

Organisational development < HEALTH SERVICES ADMINISTRATION & Keywords: MANAGEMENT, Quality in health care < HEALTH SERVICES ADMINISTRATION & MANAGEMENT, ONCOLOGY

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 25 BMJ Open

1 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on October 1, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 25

37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44

45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 25 BMJ Open

1 2 3 Processes of care and survival associated with treatment in specialist teenage and young 4 adult cancer centres: results from the BRIGHTLIGHT cohort study 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 7 Lorna A Fern PhD, Rachel M Taylor PhD, Julie A Barber PhD, Javier Alvarez Galvez PhD, 8 Richard Feltbower PhD, Sarah Lea PhD, Ana Martins PhD, Stephen Morris PhD, Louise 9 Hooker MSc, Faith Gibson PhD, Rosalind Raine PhD, Dan P Stark MD, Jeremy S Whelan MD 10 11 12 13 Lorna Fern, Teenage and Young Adult Cancer Researcher and Patient/Public Involvement 14 Lead, Cancer Division, University College London Hospitals NHS Foundation Trust, London 15 NW1 2PG 16 17 18 Rachel Taylor, DirectorFor of the peer CNMAR, Centrereview for Nurse, Midwifeonly and AHP Led Research 19 (CNMAR), University College London Hospitals NHS Foundation Trust, London NW1 2PG 20 21 Julie Barber, Associate Professor in Medical Statistics, Department of Statistical Science, 22 23 University College, London WC1E 6BT 24 25 Javier Alvarez-Galvez, Research Fellow, Department of Biomedicine, Biotechnology and 26 Public Health, University of Cadiz, Avda. Ana de Viya, 52, 11009 Cádiz, Spain 27 28 Richard Feltbower, Senior Lecturer in Epidemiology, Leeds Institute for Data Analytics, 29 30 School of Medicine, University of Leeds, Leeds, LS2 9JT. 31 32 Sarah Lea, Research Facilitator, Cancer Division, University College London Hospitals NHS 33 Foundation Trust, London NW1 2PG 34 35 36 Ana Martins, Research Associate, Cancer Division, University College London Hospitals

37 NHS Foundation Trust, London NW1 2PG http://bmjopen.bmj.com/ 38 39 Stephen Morris, Professor of Health Economics, Primary Care Unit, Department of Public 40 41 Health and Primary Care, University of Cambridge, United Kingdom 42 43 Louise Hooker, teenage Cancer Trust Lead Nurse, Wessex Teenage and Young Adult 44 Cancer Service, University Hospital Southampton, Southampton, SO16 5YA 45 on October 1, 2021 by guest. Protected copyright. 46 47 Faith Gibson, Professor of Child Health and Cancer Care, School of Health Sciences, 48 Faculty of Health and Medical Sciences, University of Surrey, Guildford, Surrey GU2 7XH 49 and Centre for Outcomes and Experience Research in Children’s Health, Illness and 50 Disability (ORCHID), Great Ormond Street Hospital for Children NHS Foundation Trust, 51 London WCIN3JH 52 53 54 Rosalind Raine, Professor of Health Services research, Department of Applied Health 55 Research, Director, NIHR CLAHRC North Thames, NIHR Senior Investigator, University 56 College London, London WC1E 7HB 57 58 59 Dan Stark, Professor of Teenage and Young Adult Cancer Research and Consultant 60 Medical Oncologist, Leeds Institute of Medical Research at St James's, Leeds LS9 7TF

1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 25

1 2 3 4 Jeremy Whelan, Professor of Cancer Medicine, Cancer Division, University College London 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 Hospitals NHS Foundation Trust, London NW1 2PG 7 8 Correspondence: Dr Rachel Taylor, [email protected] 9 10 Keywords: BRIGHTLIGHT, teenagers and young adults, cancer, observational research, 11 12 cohort, outcome, survival 13 14 Abstract = 297/300 15 Word count = 3,642 16 17 Number of Tables = 6 18 Number of Figures For= 1 peer review only 19 20 21 ABSTRACT 22 23 24 Objective: Survival gains in teenagers and young adults (TYA) are reported to be lower than 25 children and adults for some cancers. Place of care is implicated, influencing access to 26 specialist TYA professionals and research. Consequently, age-appropriate specialist cancer 27 28 care is advocated for TYA although systematic investigation of associated outcomes is 29 lacking. In England, age-appropriate care is delivered through 13 Principal Treatment 30 Centres (TYA-PTC). BRIGHTLIGHT is the national evaluation of TYA cancer services to 31 examine outcomes associated with differing places and levels of care. We aimed to examine 32 the association between exposure to TYA-PTC care, survival and documentation of clinical 33 34 processes of care. 35 36 Design: Prospective cohort study 37 http://bmjopen.bmj.com/ 38 Setting: 109 National Health Service (NHS) hospitals across England 39 40 41 Participants: 1,114 TYA, aged 13-24, newly diagnosed with cancer between 2012-2014 42 43 Intervention: Participants were assigned a TYA-PTC category dependent on the proportion 44

2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 25 BMJ Open

3 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 25

4 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 25 BMJ Open

1 2 3 care in a children’s or adult cancer unit (SOME-TYA-PTC). Sample size calculations were 4 based on the primary outcome measure of the cohort, quality of life (18). 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 7 Participants and setting 8 The BRIGHTLIGHT cohort comprised young people aged 13-24 years, newly diagnosed 9 with cancer (ICD-10 codes C00-C97) in an English hospital and recruited within four-months 10 of diagnosis. Eligibility criteria were as inclusive as possible so no restriction according to 11 12 language or any sensory impairment that affected communication was applied. The only 13 exclusion criteria were: young people receiving a custodial sentence; if the young person 14 was not anticipated to be alive at the first point of data collection (6-months after diagnosis); 15 recurrence of a previous cancer or they were not capable of completing a survey, e.g. 16 17 sedated and in intensive care. The processes for recruitment are reported in detail 18 elsewhere (18, 20, For21). BRIGHTLIGHT peer wasreview open to recruitment only in 109 NHS hospitals in 19 England, of whom 97 hospitals recruited at least one young person. Young people were 20 recruited between October 2012 and April 2015 (diagnosed between July 2012 and 21 December 2014). They gave written consent (parent consent also obtained from those less 22 23 than 16 years); the study was approved by London-Bloomsbury NHS Research Ethics 24 Committee and the Confidentiality Advisory Group. 25 26 Data collection 27 We assessed documentation of the following clinical processes: 28 29 30  Histological diagnosis 31  Molecular confirmation of diagnosis (where relevant) 32  Cancer stage or prognostic group defined (for leukaemia, total white blood cell count) 33 34  Initial treatment plan 35  Treatment protocol for systemic therapy and/or for radiotherapy 36  Evidence of MDT communication including children’s, TYA or site specific 37 http://bmjopen.bmj.com/ 38  Assessment by supportive care services based on evidence in notes of contact with 39 clinical nurse specialist plus one other member of the MDT (social worker, youth 40 support coordinator, counsellor, psychologist, dietician, physiotherapist, occupational 41 therapist), 42  A record of fertility being discussed 43 44  A record of consideration for inclusion in a clinical trial

5 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 25

1 2 3 Analysis was based on a predefined analysis plan using STATA Version 15. For each 4 clinical process outcome, the proportion of patients where the item was found to be 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 documented in clinical records was reported by category of care. Proportions were 7 compared across groups using Chi squared tests (including a trend test). 8 9 Survival time for each participant was calculated from date of diagnosis to date of death or 10 censored at the date last known to be alive up to 29th October 2018. Kaplan Meier survival 11 12 curves were plotted for each category of care (NO-TYA-PTC, SOME-TYA-PTC and ALL- 13 TYA-PTC) and estimates of cumulative survival at 1 to 4 years (with 95% confidence 14 intervals) were calculated. The relationship between survival time and TYA category was 15 investigated using a Cox regression model adjusted for confounding factors identified using 16 17 a causal inference approach and based on the conceptual model underpinning the 18 BRIGHTLIGHT SurveyFor (19) inpeer the form ofreview a Directed Acyclic only Graph (DAG) (Supplemental 19 file; DAGitty software www.dagitty.net). Factors adjusted for were age at diagnosis, sex, type 20 of cancer (leukaemia, lymphoma, brain and central nervous system, bone tumours, 21 sarcoma, germ cell, melanoma, carcinomas, other), socioeconomic status (Index of Multiple 22 23 Deprivation rank (IMD) (23)), severity of cancer (least, intermediate, most (18)), ethnicity 24 (white, other), days in hospital over 12 months since diagnosis, treatment received in 12 25 months since diagnosis (Systemic anticancer therapy (SACT) only, radiotherapy (RT) only, 26 Surgery only, Surgery & SACT/SACT & RT & Surgery/RT & SACT/Surgery & RT/Transplant, 27 Other). Geographical region of treatment (North East, North West, Yorkshire, East 28 29 Midlands, West Midlands, London, South East and South West) was included in the model 30 as a random effect (frailty term). The proportional hazards assumptions of the Cox 31 regression model were checked. Models were extended to include interaction terms to 32 investigate whether the association between TYA group and survival was different by age at 33 34 diagnosis (using categories of 13 -18 and 19 - 24 years, and age in years) and tumour type 35 (using categories, haematology and oncology). 36

6 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from

Page 9 of 25 BMJ Open

BMJ Open Page 10 of 25

Page 11 of 25 BMJ Open

45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

10 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from

Page 13 of 25 BMJ Open

BMJ Open Page 14 of 25

16 http://bmjopen.bmj.com/ 17 18 19 20 21 22 23

1 2 3 Survival 4 The duration of follow-up by October 2018 is shown in Table 3. The number of deaths in the 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 NO-TYA-PTC group was 27 (8%), compared to 35 (13%) in ALL-TYA-PTC and 91 (22%) in 7 SOME-TYA-PTC. The cumulative probability of survival by time since diagnosis for the TYA- 8 PTC categories is shown in Figure 1 and Table 4. Although survival probabilities at one year 9 were similar there was clear divergence between the groups over the following time period, 10 such that probabilities were highest for those receiving NO-TYA-PTC, followed by ALL-TYA- 11 12 PTC care, then SOME-TYA-PTC care. Following full adjustment for confounding factors, 13 regression (Table 5) showed there was no evidence of a relationship between the category of 14 care and hazard (risk) of death. 15 16 17 Table 3: Duration of follow-up 18 ForNO-TYA-PTC peerSOME-TYA-PTC review onlyALL-TYA-PTC TOTAL 19 N=359 N=415 N=270 N=114 20 Median (IQR) 1839 1743 1747 1779 21 follow up (days) (1597 to 2041) (1474 to 1991) (1536 to 2023) (1536 to 2023) 22 23 24 25 Table 4: Estimated cumulative survival probabilities by categories of TYA care and year 26 from diagnosis (95% CI) 27 28 NO-TYA-PTC SOME-TYA-PTC ALL-TYA-PTC 29 1 year 0.98 (0.96 to 0.99) 0.97 (0.95 to 0.99) 0.98 (0.95 to 0.99) 30 2 years 0.95 (0.92 to 0.97) 0.89 (0.86 to 0.92) 0.93 (0.89 to 0.95) 31 3 years 0.94 (0.91 to 0.96) 0.83 (0.79 to 0.86) 0.90 (0.85 to 0.93) 32 33 4 years 0.93 (0.90 to 0.95) 0.80 (0.76 to 0.84) 0.89 (0.84 to 0.92) 34 35 36

13 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 25

14 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 25 BMJ Open

1 2 3 be associated with specialist age-appropriate care (16,17), therefore it is not surprising these 4 appear to be more frequently documented in the ALL and SOME group. 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 Our results suggest differences between the groups in these measures of the quality of 7 cancer care delivered to young people. However, this did not appear to have an impact on 8 survival outcomes. One-year survival was similar between the groups; at four years, survival 9 10 was highest in the NO-TYA-PTC group, followed by ALL-TYA-PTC and lowest in the SOME- 11 TYA-PTC group. However, this was not significant. One of the reasons for implementing a 12 new model of care in 2005 specific for TYA with cancer was due to the disparity in survival 13 compared to children and older adults (12). It is therefore disappointing that there were no 14 survival differences noted between the three categories. Alternatively, we could view this as 15 a positive finding - wherever young people choose to be treated, their survival outcomes are 16 the same. The work we did with young people to develop BRIGHTLIGHT highlighted that 17 they did not perceive survival alone as the most import outcome. Quality of life and the ability 18 For peer review only 19 to get back on with life were as important (18,19). While we have shown a better 20 improvement in quality of life when treated in a TYA-PTC (23), we have yet to ascertain 21 whether young people's reintegration into life when treatment ends is also better. 22 23 There was weak evidence that increasing age was associated with higher risk of death for 24 those in the SOME-TYA-PTC compared to NO-TYA-PTC and ALL-TYA-PTC, an important 25 finding given that it is this group who have a choice over where to receive their care (12). We 26 believe that further investigation into the lower survival in the SOME-TYA-PTC group is 27 warranted, particularly the association with age. Those aged over 19 years at diagnosis had 28 the more pronounced effects in the subgroup analysis, but the direction of the effect differed 29 30 substantially for under 19s compared to the overall effect. 31 There is a paucity of existing literature to compare our results with and comparisons are 32 33 further confounded by variation in healthcare systems, distinct models of specialist age- 34 appropriate care adopted and the international definition of TYA which can extend up to 39 35 years in some countries (24). A previous retrospective regional study of children and TYA in 36 England found a survival benefit of being treated in a PTC for poor prognosis leukaemia and 37 a converse relationship for those with soft tissue sarcoma, no significant differences in http://bmjopen.bmj.com/ 38 survival were observed for those with lymphoma, CNS, bone and germ cell tumours (25). Of 39 note, a previous study has also shown those receiving ‘SOME’ specialist care have poorer 40 survival for some indications ( Birch 2013, unpublished thesis). These studies suggest that 41 42 some tumour groups may benefit from care at the PTC however, due to our previously 43 reported difficulties with recruitment and reduction in sample size (20) we were unable to 44 conduct the detailed analysis of individual cancer types as planned, and thus benefits of the 45 PTC may be masked within the grouping of ‘haematology’ and ‘solid tumours’. on October 1, 2021 by guest. Protected copyright. 46 47 Limitations 48 49 Despite our study including a large, broadly representative sample of newly diagnosed TYA 50 with cancer followed-up for 3 years, and our analyses being adjusted for factors known to 51 affect outcomes in cancer there are some limitations to our study. Our definition of 52 ‘specialist’ was based on the TYA-PTC care model as defined by the NICE Improving 53 Outcomes Guidance Issued in 2005 (12), which does not necessarily reflect current delivery 54 of age-appropriate care (9). The study population were recruited during a period of evolution 55 of TYA services in England, therefore the models of care are unlikely to reflect current 56 57 practice, particularly as we have identified that specialist age-appropriate care takes time to 58 develop (13). Additionally, categorising TYA-PTC assumes that all PTCs are equal and does 59 not measure the quality of care delivered. We know national variation exists in configuration 60 and maturity of services (26), particularly during 2012-2014 when patients were recruited.

15 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 25

1 2 3 Further, due to the coding of hospital inpatient data it is possible that some patients have 4 been misattributed as receiving care in the TYA-PTC when they may have been cared for in 5 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from a Trust which had a TYA-PTC but care was delivered at a different hospital and not in the 6 7 TYA unit. An additional limitation of the categorisation of care was that it was based on 8 previous work (Birch 2013, unpublished thesis), which only included inpatient admission 9 data. Potentially, hospital visits involving treatment as outpatient care were not included. 10 This could have resulted in patients been misclassified as ALL-TYA-PTC or NO-TYA-PTC. 11 12 Consideration must also be given to additional factors influencing survival outcomes which 13 we did not collect or measure. These include deviation from the intended treatment plan 14 such as the proportion of treatment delivered, delays/reduction in delivery and toxicity. 15 Therefore, it was not possible to determine the dose or type of chemotherapy or 16 radiotherapy received by patients in each group and these would be important determinants 17 of survival. Our scale for place of care was derived only from inpatient care and not care or 18 For peer review only 19 treatment delivered as an outpatient. Thus, we may have missed considerable elements of 20 the care received. 21 22 Further to this, overall survival of the BRIGHTLIGHT cohort was lower than those diagnosed 23 during the same period but not recruited, therefore our findings may not reflect the 24 experience of the whole TYA population (18). We also do not know the decision-making 25 processes behind referral of patients into each TYA-PTC group at diagnosis, it may be that 26 patients with better prognosis are treated more locally with site specific expertise competent 27 at treating the cancer with good survival outcomes, while those with more complex disease 28 and holistic needs are referred into the specialist TYA service. 29 30 Conclusion 31 32 We have reported on the first systematic longitudinal evaluation of cancer services for young 33 people. Young people were more likely to have had documentation of access to supportive 34 care services or have a fertility discussion if they had some or all of their care delivered by 35 the TYA-PTC, which existing literature supports as important for TYA. 36 37 Overall, survival at four years was good across all three categories of care with some http://bmjopen.bmj.com/ 38 differences between the NONE, ALL and SOME groups as defined by NICE improving 39 40 outcomes service specification in 2005. The factors contributing to survival differences 41 between the groups warrants further investigation particularly the relationship between 42 survival, level of TYA care and age. BRIGHTLIGHT results are immediately important for 43 current healthcare recommendations for young people with cancer in England. The currently 44 proposed model of care proposed by NHS England advocates ‘Joint Care’ but with an 45 emphasis to increase communication between the TYA-PTCs and selected local hospitals. on October 1, 2021 by guest. Protected copyright. 46 Further enquiry is required with additional prospective data collection to assess whether this 47 proposed ‘Joint Care’ would generate a similar pattern of survival trends as the ‘SOME-TYA- 48 49 PTC’ Group in our study. 50 51 52 REFERENCES 53 54 1. Birch JM, Alston RD, Kelsey AM, Quinn MJ, Babb P, McNally RJQ. Classification and 55 incidence of cancers in adolescents and young adults in England 1979–1997. British Journal 56 of Cancer. 2002;87(11):1267-74. 57 2. Fern LA, Birch R, Whelan J, Cooke M, Sutton S, Neal RD, et al. Why can't we 58 improve the timeliness of cancer diagnosis in children, teenagers, and young adults? BMJ. 59 2013;347:f6493. 60

16 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 25 BMJ Open

1 2 3 3. Fern LA, Lewandowski JA, Coxon KM, Whelan J, National Cancer Research Institute 4 T, Young Adult Clinical Studies Group UK. Available, accessible, aware, appropriate, and 5 acceptable: a strategy to improve participation of teenagers and young adults in cancer BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 trials. Lancet Oncol. 2014;15(8):e341-50. 7 4. Barr RD, Ries LA, Lewis DR, Harlan LC, Keegan TH, Pollock BH, et al. Incidence 8 and incidence trends of the most frequent cancers in adolescent and young adult Americans, 9 10 including "nonmalignant/noninvasive" tumors. Cancer. 2016;122(7):1000-8. 11 5. Tricoli JV, Blair DG, Anders CK, Bleyer WA, Boardman LA, Khan J, et al. Biologic 12 and clinical characteristics of adolescent and young adult cancers: Acute lymphoblastic 13 leukemia, colorectal cancer, breast cancer, melanoma, and sarcoma. Cancer. 14 2016;122(7):1017-28. 15 6. Herbert A, Lyratzopoulos G, Whelan J, Taylor RM, Barber J, Gibson F, et al. 16 Diagnostic timeliness in adolescents and young adults with cancer: a cross-sectional 17 analysis of the BRIGHTLIGHT cohort. Lancet Child Adolesc Health. 2018;2(3):180-90. 18 7. Boissel N, BaruchelFor A.peer Acute lymphoblastic review leukemia only in adolescent and young adults: 19 treat as adults or as children? Blood. 2018;132(4):351-61. 20 8. Fern LA, Taylor RM, Whelan J, Pearce S, Grew T, Brooman K, et al. The art of age- 21 appropriate care: reflecting on a conceptual model of the cancer experience for teenagers 22 and young adults. Cancer Nurs. 2013;36(5):E27-38. 23 9. Lea S, Taylor RM, Martins A, Fern LA, Whelan JS, Gibson F. Conceptualizing age- 24 appropriate care for teenagers and young adults with cancer: a qualitative mixed-methods 25 study. Adolesc Health Med Ther. 2018;9:149-66. 26 10. Barr RD. Planning a Comprehensive Program in Adolescent and Young Adult 27 Oncology-A Collision with Reality. J Adolesc Young Adult Oncol. 2016;5(4):303-9. 28 11. Morgan S, Davies S, Palmer S, Plaster M. Sex, Drugs, and Rock ‘n’ Roll: Caring for 29 30 Adolescents and Young Adults With Cancer. Journal of Clinical Oncology. 31 2010;28(32):4825-30. 32 12. Excellence NIfC. Improving Outcomes Guidance for Children and Young People with 33 Cancer. 2005. 34 13. Stiller CA. Centralisation of treatment and survival rates for cancer. Arch Dis Child. 35 1988;63(1):23-30. 36 14. Stiller CA. Centralised treatment, entry to trials and survival. Br J Cancer.

37 1994;70(2):352-62. http://bmjopen.bmj.com/ 38 15. Greenberg M, Klassen A, Gafni A, McBride ML, Albritton K. Outcomes and metrics: 39 measuring the impact of a comprehensive adolescent and young adult cancer program. 40 Cancer. 2011;117(10 Suppl):2342-50. 41 16. Rae CS, Pole JD, Gupta S, Digout C, Szwajcer D, Flanders A, et al. Development of 42 System Performance Indicators for Adolescent and Young Adult Cancer Care and Control in 43 Canada. Value Health. 2020;23(1):74-88. 44 17. Sironi G, Barr RD, Ferrari A. Models of Care-There Is More Than One Way to on October 1, 2021 by guest. Protected copyright. 45 Deliver. Cancer J. 2018;24(6):315-20. 46 18. Taylor RM, Fern LA, Barber J, Alvarez-Galvez J, Feltbower R, Morris S, et al. 47 Description of the BRIGHTLIGHT cohort: the evaluation of teenage and young adult cancer 48 49 services in England. BMJ Open. 2019;9(4):e027797. 50 19. Taylor RM, Fern LA, Solanki A, Hooker L, Carluccio A, Pye J, et al. Development and 51 validation of the BRIGHTLIGHT Survey, a patient-reported experience measure for young 52 people with cancer. Health Qual Life Outcomes. 2015;13:107. 53 20. Kenten C, Martins A, Fern LA, Gibson F, Lea S, Ngwenya N, et al. Qualitative study 54 to understand the barriers to recruiting young people with cancer to BRIGHTLIGHT: a 55 national cohort study in England. BMJ Open. 2017;7(11):e018291. 56 21. Taylor RM, Fern LA, Aslam N, Whelan JS. Direct access to potential research 57 participants for a cohort study using a confidentiality waiver included in UK National Health 58 Service legal statutes. BMJ Open. 2016;6(8):e011847. 59 60

17 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 25

1 2 3 22. Husson O, Zebrack BJ, Block R, Embry L, Aguilar C, Hayes-Lattin B, et al. Health- 4 Related Quality of Life in Adolescent and Young Adult Patients With Cancer: A Longitudinal 5 Study. J Clin Oncol. 2017;35(6):652-9. BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 6 23. Taylor RM, Fern LA, Barber J, Alvarez-Galvez J, Feltbower R, Lea S, et al. 7 Longitudinal cohort study of the impact of specialist cancer services for teenagers and young 8 adults on quality of life: outcomes from the BRIGHTLIGHT study. BMJ Open. 9 10 2020;10(11):e038471. 11 24. Pini SA, Gibson F, Fern LA, Morgan SJ, Phillips RS, Stark DP. Multi-Professional 12 Perspectives on Adolescent and Young Adult Oncology Across Europe: An e-Delphi Survey. 13 J Adolesc Young Adult Oncol. 2017;6(1):178-85. 14 25. Fairley L, Stark DP, Yeomanson D, Kinsey SE, Glaser AW, Picton SV, et al. Access 15 to principal treatment centres and survival rates for children and young people with cancer in 16 Yorkshire, UK. BMC Cancer. 2017;17(1):168. 17 26. Vindrola-Padros C, Taylor RM, Lea S, Hooker L, Pearce S, Whelan J, et al. Mapping 18 Adolescent CancerFor Services: peer How Do Young review People, Their only Families, and Staff Describe 19 Specialized Cancer Care in England? Cancer Nurs. 2016;39(5):358-66. 20 21 Author contributions: 22 RMT, LAF, JB, DS, SM, RF, LH, FG, RR, JSW were involved in developing the protocol. 23 24 RMT coordinated the running of the study and was responsible for data acquisition. JB, JGA, 25 RMT, LAF, AM, SL, SM, RF, DS, JSW contributed to the analysis. RMT, LAF, JB and JSW 26 drafted the manuscript. All authors critically revised and approved the final manuscript. 27 28 29 Patient and public involvement: 30 Young people have been involved in this study from the feasibility stage onward. They were 31 involved in study development, acted as co-researchers and were instrumental in the design 32 and methods of the study. A representative of the Young Advisory Panel (YAP) was a co- 33 applicant on the grant and the YAP have been part of the management of the study since 34 35 the grant was awarded in 2011. Details of the extent of young people’s involvement in 36 BRIGHTLIGHT is provided in reference 13.

37 http://bmjopen.bmj.com/ 38 Acknowledgements: 39 We would like to thank the members of our Young Advisory Panel (Zeena Beale, Ciaran 40 41 Fenton, Emily Freemantle, Jaasjan Guvindia, Laura Haddard, Steph Hammersley, Amy 42 Lang, Joshua Lerner, Tanya Loughlin, Jason, Sin Jin Loo, Jennifer Miller, Maria Onasanya, 43 Arif Nasir, Steph Still, Poppy Richards, Amy Riley, Lara Veitch, Freya Voss, JJ Wheeler, 44 Max Willliamson, Antonia Young), the 1,114 young people who consented to participate in 45 on October 1, 2021 by guest. Protected copyright. 46 BRIGHTLIGHT, healthcare professionals who approached and consented young people, 47 and ex-members of the team who have contributed to study management (Catherine 48 O’Hara, Anita Solanki, Natasha Aslam, Zuwena Fox). 49 50 We would like to dedicate this manuscript in memory of Mr Stephen Sutton and Mr Mathew 51 52 Cook who were instrumental to study set up, design and management. Both of whom died 53 from their cancer during the study. 54 55 We would also like to thanks the following for all their support with recruitment to 56 BRIGHTLIGHT: the National Cancer Research Institute, especially Dr Eileen Loucaides and 57 58 the Secretariat; Matt Seymour, Matt Cooper and Karen Poole at the former National Cancer 59 Research Network; Maria Khan and Sabrina Sandhu (North West Knowledge Intelligence 60

18 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 25 BMJ Open

19 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 25

45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

20 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 25 BMJ Open

37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44

45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

21 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 25

1 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 Figure 1 30 31 75x52mm (300 x 300 DPI) 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on October 1, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from

Page 25 of 25 BMJ Open

1 2 3 Supplemental file 4

5 6 Figure A1: Matrix representing the causal diagram explaining confounding variable to enter in the Directed Acyclic Graph (DAG) 7 TYA QOL Survival Age Gender CS CT SES Ethnicity Geography DoH Treatment RtD 8 CoC 9 TYA CoC         O    10 Survival            11 12 Age For O peer  review O O only    13 Gender   O O O     14 CS         15 CT       

16 http://bmjopen.bmj.com/ 17 SES       18 Ethnicity      19 Geography     20 DoH 21    22 Treatment   23 RtD 

43 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2020-044854 on 7 April 2021. Downloaded from

BMJ Open Page 26 of 25

Page 27 of 25 BMJ Open