Comparison of healthcare access for the general population in Yorkshire under the age of 65 to a population who were diagnosed with childhood or young adult cancer

The University of Leeds require data specifically for comparison of a cohort within the Yorkshire Specialist Register of Cancer in Children and Young People (YSRCCYP), with the general population of Yorkshire.

The YSRCCYP was originally set up in collaboration with local clinicians to provide research information. Since 1994, the YSRCCYP database and research programme has been managed by the University of Leeds’ Division of Epidemiology & Biostatistics. The University of Leeds is the Data Controller for the YSRCCYP with sole responsibility for determining the purposes for which and the manner in which any personal data are processed. The work is currently funded solely by the Candlelighters Trust. The Candlelighters Trust do not determine how data is collated or processed for the study and will only access data that is aggregated with small numbers suppressed.

The YSRCCYP is a regional population based register containing detailed, record-level demographic and clinical information on children and young adults aged 0-29 years diagnosed with cancer and benign central nervous system tumours since 1974. The YSRCCYP covers the Yorkshire and Humber Strategic Health Authority (SHA) which has a total population of 5 million people. The purpose of the YSRCCYP is to facilitate population-based epidemiological and health services research.

Historically, cancer registration records have been identified primarily through hospital records involving manual abstraction of information from hospital notes. In recent years, the data controller has established a secure, electronic feed of information from local (within NHS Trusts across Yorkshire) patient pathology management systems (PPM), the National Cancer Registration and Analysis Service (NCRAS), as well as primary care datasets, in order to improve the efficiency of the Yorkshire Register data collection. The flow of data from local trusts into the registry is covered under section 251 approval from the HRA CAG to address the common law duty of confidence. Section 251 support is not applicable to this Agreement as only the flow of pseudonymised data is permitted.

From October 2021 NCRAS became part of the National Disease Registration Service (NDRS) service managed by NHS Digital. Data continues to flow to University of Leeds from the NDRS service, however this is a separate flow of data to the data disseminated under this Agreement, and the two sets of data are not linked in any way.

Where data on cancer diagnosis and treatment are missing or incomplete from the electronic data sources, the study team manually abstract information from local hospital notes and patient management systems. The data comprises personal and demographic information (e.g. name, date of birth, address, postcode) along with diagnostic and clinical data on treatment (chemotherapy drugs and doses, radiotherapy site and dose, surgery). More recently, enhanced treatment information on chemotherapy and radiotherapy will be obtained through NCRAS, including the national Systemic Anti-Cancer Therapy datasets. All data provided form third-party sources are added to the registry at a pseudonymised level. The data provided from NHS Digital is not added to the registry, comparison on participants is made using only a pseudonymised patient ID. No attempt is made to re-identify a participant using either the data from NHS Digital or any data within the registry.

Data on 12,000 patients have been collected from 1974 onwards, however linked HES and mental health data was required for only 9,000, as the cohort shared with NHS Digital excludes participants who passed away before 1996. The cohort submission under this Agreement will consist of approximately 9,000 patients.

The YSRCCYP is a population-based register; inclusion is based on the patient's residential address at diagnosis, those resident in Yorkshire & Humber SHA. Any individuals being treated in Yorkshire hospitals but resident outside the region will be excluded. Those on the register who move away from the Region remain included in case of any follow-up treatment or care. All patient information is retained beyond the age of 30 years for the purpose of tracking their long-term cancer outcomes. A previous extract relating to 5-37 year olds admitted between 2011-2016 comprised over 11 million records and 3.5 million unique HES IDs. University of Leeds are requesting extended extracts for 5-64 year olds so the number of records is likely to be higher than that. Until the data is produced an exact figure cannot be determined.

The sole data controller is the University of Leeds. The legal basis for dealing with people’s personal data for research data, is ‘task in the public interest’, or in other words, the University are collecting data that they need in order to complete their research, which aims to improve healthcare for the general public. The justification for processing the data by University of Leeds is Article 6 (1)(e) of the GDPR: (processing is necessary for the performance of a task in the public interest or in the exercise of official authority vested in the controller). The justification for the processing of the special category data (health data) by University of Leeds is Article 9 (2)(j) of the GDPR: (processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes in accordance with Article 89(1) based on Union or Member State law which shall be proportionate to the aim pursued, respect the essence of the right to data protection and provide for suitable and specific measures to safeguard the fundamental rights and the interests of the data subject.

Patients are made aware of all data flows through leaflets distributed at the main Principal Treatment Centres, posters on hospital wards, webinars with the local charity funders (Candlelighters Trust) and the YSRCCYP register website and fair processing statement which is published on the University of Leeds website.

The data controller involves members of the public and participants through presentations of results. Feedback from presentations researchers have made to families and cancer survivors, have shown that they are hugely grateful for the Register and research carried out on their behalf. Those attending these presentations felt that they had been sufficiently informed about the research, and are comfortable and supportive of the need to collect and process this data without informed consent.

The YSRCCYP research database has its' own Scientific Advisory Group which includes representation from the following areas – university research, public health, clinical medicine, health care professionals, patients and lay persons. The Advisory Group is responsible for ensuring that appropriate procedures are in place to ensure the physical security of the data and its release. The Advisory group do not access or determine how the data under this Agreement will be used and are therefore not deemed to be data controllers.

Data requested are record level Hospital Episode Statistics (HES) and Mental Health data which will be used by NHS Digital to create a cohort consisting of inpatient hospital admissions and mental health episodes for the general population in Yorkshire under the age of 65.

This comparison cohort data will be used to compare against data on a population of the same age range within the YSRCCYP cohort, who were diagnosed with childhood or young adult cancer (derived from data provided under a separate Data Sharing Agreement (reference: NIC-11809-H1Y3W)). Data are only requested for the period from 2006/07 to 2021/22 (or latest available data).

Under a previous iteration of this Agreement the University of Leeds received tabulated extracts of mental health data, including the Mental Health and Learning Disabilities Dataset and the Mental Health Minimum Data Set - covering the period from 2006/07 to 2015/16, aggregated with small number suppressed. This data was broken down by age and single year, rather than record-level data and given the age of the cohort at that time, was restricted to patients who were aged 46 or less. However, this did not provide sufficient information regarding mental health in order to calculate hospitalisation rate ratios, or adequately adjust for the effects of or stratify by gender, deprivation, ethnicity and age. University of Leeds also previously received HES data, aggregated with small numbers unsuppressed, for the period 1996/97 - 2015/16. Again, this data was found to be insufficient to meet the requirements for the study.

Therefore University of Leeds is now requesting pseudonymised, record level HES and Mental Health data for the period between 2006/07 and 2021/22, and, given the current age of cohort population, the data required moving forward will now include all patients aged 65 or younger.

The HES and mental health data are not added into the YSRCCYP research database. The two datasets are stored separately but contain common unique study IDs enabling data to be compared at record level. Where required for specific research, relevant data are extracted from the respective databases, compared and analysed by the YSRCCYP research team.

The data under this Agreement will be compared with (but not linked with) data on inpatient hospital admissions and Mental Health data for the YSRCCYP cohort who were diagnosed with childhood or young adult cancer. The aim is to assess whether certain hospital admissions or mental health episodes are more (or less) common amongst a population of survivors of childhood and young adult cancers following treatment, compared to the general population. The risk of inpatient admissions and mental health episodes of a certain diagnosis in the cancer population will also be compared. The YSRCCYP research team aims to look at the whole admission pattern of patients, not simply those that occur in the primary diagnosis fields and therefore require an episode level extract as opposed to aggregated counts of admission.

The YSRCCYP research team requires HES data for the full periods of 1996/97 to 2021/22 (or latest available data) for several reasons. Firstly in order to address aim 1 and 4-6 investigating changes in levels of hospitalisation over time and within specific cancer types, age groups and sociodemographic groups. Secondly to maximise statistical power for analyses given the rarity of childhood and young adult cancer; also to assess changes in access over time to specialist cancer care services, as NHS policy regarding the recommendations for treatment of children and young people with cancer has changed with the opening of Principal Treatment Centres in hospitals across in England throughout this time frame.

Please note the objectives listed below apply to both this Agreement and to NIC-11809-H1Y3W because it is only when using data from both Agreements that the below objectives can be achieved.

The YSRCCYP research team’s research plans include the following objectives:

1) To describe the total burden of physical and mental health hospitalisation among the Yorkshire cancer population aged 0-29 years, to identify clinical and sociodemographic factors which influence the likelihood of hospitalisation and to investigate how hospitalisation rates have changed since 1997.

2) To understand patient care pathways (the specific route that a particular patient takes) through the NHS before, during and after cancer diagnosis. This includes assessment of time to diagnosis for children and young adults diagnosed with cancer under the age of 30 years to identify where improvements can be made to minimise delays in diagnosis leading to better prognosis and less stress and anxiety on patients and their families.

3) To calculate the risks and costs to the NHS of adverse physical and mental health events requiring hospital admission for survivors of cancer in this age group so that clinicians can provide appropriate follow-up care.

4) To identify the risks of cardio-metabolic disease in long-term childhood and young adult cancer survivors. Specific risk markers for cardiovascular disease plus metabolic factors associated with metabolic syndrome and type II diabetes will be collected. The study team will also obtain a range of additional biomarkers of cardiovascular risk. This data will facilitate a programme of research assessing the prevalence of endocrine complications and cardiometabolic late effects in long-term childhood and young adult cancer survivors.

5)i)To identify the impact of cancer treatment on kidney toxicity and mental health, specifically anxiety and depression. In this project researchers aim to enhance the treatment data held in the register through linkage with the national Systemic Anti-Cancer Therapy (SACT) dataset and hospital electronic prescribing systems such as ChemoCare*. This will enable researchers to compare the chemotherapy doses and intensities given to patients with the same tumour types and see if this has any effect on outcomes including survival and relapse. Researchers will also be able to look at the toxicities experienced by patients by reviewing routine measurements and blood tests collected prior to chemotherapy. From this, researchers will be able to see if any dose modifications were made as a result and again look at whether this had any effect on patient outcome. By comparing patients treated at different hospitals. researchers will, for some tumour types, be able to see whether, and if so how, practices differ between Principal Treatment Centres and peripheral hospitals.

ii) An extension of this project is to look in more detail at the impact of cancer treatments on kidney toxicity. Routine bloods tests and measurements held will be enhanced by the collection of urinary analyses, microbiology results, radioisotope measurements and the prescription of any anti-hypertensive medications. These additional fields will give researchers a greater insight into the kidney toxicity caused.

*Please note, the data added to the YSRCCYP from data sources such as ChemoCare, will never be linked to NHS Digital at record level. Only aggregated outputs with small numbers suppressed will be compared.

6) To provide contextual information on existing physical and mental health morbidity when evaluating educational and employment outcomes. As part of the ongoing research the study team collect additional information for patients on the register attending the long term follow-up clinics at Leeds Teaching Hospitals Trust. Patients attending these clinics complete a holistic needs assessment which includes the completion of the distress thermometer and a problems check list. These data items are used as a measure of psychological health to assess the prevalence of distress in long term cancer survivors and the associations between patient characteristics and levels of distress.

To address aims 1 and 4, the YSRCCYP research team will utilise HES and Mental Health data to investigate long term risks of all major morbidity (e.g. cardio-metabolic, respiratory, mental health illness) in the cohort and identify sociodemographic and clinical factors which may affect these risks. The YSRCCYP research team also wish to determine the relative excess risk of these conditions within the cancer cohort compared to the general background population and, in order to make this comparison, requires a separate pseudonymised extract of HES data containing all episodes for YSRCCYP members in the Yorkshire and Humber SHA area only under the age of 65 at admission (the oldest person currently registered in the database). This separate extract is covered under this DSA.

To address aims 2 and 3 evaluating clinical care pathways and time to diagnosis also require data of all admissions prior to cancer diagnosis. The late health effects for childhood and young adult cancer survivors may occur any time after treatment ends and the risk of late effects increases as the cohort ages. In order to fully evaluate the total burden of adverse health events in these survivors’, data are required for as long a time period as possible. This may also include any hospital admissions prior to the patient’s cancer diagnosis to identify any underlying health conditions. The YSRCCYP research team is also notified about any subsequent malignant neoplasms from the National Cancer Registration and Analysis Service prospectively following the original cancer diagnosis and therefore need to retain all historical HES and mental health data in order to scrutinise any such individual’s history of hospital admissions and understand potential reasons for those who experience multiple tumour diagnoses.

Rates of admission within the cancer survivors have previously been compared to pseudonymised hospital admission rates to work out standardised hospitalization admission ratios and assess whether these differed according to cancer diagnosis, treatment, ethnic group, gender, age group, period of diagnosis and socioeconomic status, using statistical models adjusting for patient case-mix while also incorporating the general background hospital admission rates. To address aim 3 this process will be repeated using the latest pseudonymised HES extract with a focus on specific disease groups, including cardio-metabolic, kidney disease, anxiety and depression, as well as total physical and mental health morbidity data, using a similar methodology as the YSRCCYP research team’s previously published work on cardiovascular disease, respiratory morbidity and cumulative burden.

Aim 4 will examine the lifelong risk for the development of endocrine complications or adverse cardio-metabolic health outcomes attributable to the cancer and/or treatment with a specific focus on these risks in relation to ethnic group and socio-economic status.

Aim 5 will enable the YSRCCYP research team to estimate the risk of kidney disease, anxiety and depression among the Yorkshire survivorship cohort, whether this risk has changed over time and the points at which it may appear in the survivor's cancer treatment. The key clinical factors which influence the risk of kidney disease, anxiety and depression will also be identified.

A current research focus is on hospital burden around the time of diagnosis and treatment and monitoring long term risks of hospitalisation associated with cancer treatment. One specific processing activity will relate to describing the risks and prevalence of mental health illness within the cancer cohort compared to the general population. Using data from March 2020 onwards, the study team will examine any changes in the long-term health risks by identifying those individuals in the cohort who tested positive for Sars-Cov-2 infection.

Aim 6 will provide important clinical information on existing health problems, identified from HES data, when describing educational attainment and employment trajectories for the young people diagnosed with cancer in Yorkshire. Specifically, those with existing health problems will be identified and this information will be taken into account as a potential confounding factor when examining the risks of poor educational or employment outcomes among the entire survivor cohort. Furthermore, we will be able to determine whether the risks of poor educational or employment outcomes are exacerbated for those with pre-existing illness.

As a result of the Covid-19 pandemic, a patient’s Covid-19 status and related data (such as date positive status was confirmed and shielding status) are also collected. This enables an analysis of the impact of Covid-19 on cancer treatments for children and young people (such as delays in treatment or non-start of treatments) and long-term health and social outcomes. Also resulting from the changes to working practices in response to the Covid-19 pandemic, data collection is completed remotely where possible. Approval has been obtained from NHS Trusts for remote access to digitalised patient notes.

This type of epidemiological and health services research has the potential to benefit future patients by identifying risk factors which can be used by health professionals to identify those at greatest risk of mental health illness so that interventions and appropriate support can be implemented. It may also reveal important environmental risk factors, examine changes in incidence rates which may help to identify possible causes and understand survival patterns according to ethnic group and socio-economic status in order to ensure that there are no inequalities in outcomes or access to specialist cancer care for certain sub-populations. The research will also have the potential to determine the impact of the COVID-19 pandemic on long-term health outcomes since March 2020.

The YSRCCYP research team require some data items classed as sensitive. These are the Referrer code, which indicates the manner in which the patient was referred to hospital by ascertaining the code of the referring organisation. This allows the YSRCCYP research team to identify particular patient pathways which are associated with an optimal time to diagnosis, a key indicator known to influence survival. Additionally the Consultant code data field is required because it enables the YSRCYYP research team to work out whether patients receive care at specialist cancer centres as opposed to general district hospitals, in order to address important health services research questions such as: ‘Does specialist care improve patient outcomes for children and young people including length of hospital stay and reduce subsequent morbidity and mortality?’. There are currently no databases which link consultant codes to specialist cancer centres for childhood and young adult cancer, so this process needs to be done manually using cohort linked NHS Digital data and the YSRCCYP database.

UofL are keen to access data by gender, deprivation, ethnicity (South Asian, Black, White, other) as well as age to allow them to draw accurate comparisons with the cancer cohort.

The pseudonymised HES extract is used to calculate the admission rates (per 100,000 per year) for the same HES diagnoses as the patient cohort. This work has been completed for cardiovascular diseases (van Laar et al, British Journal of Cancer 2014), respiratory morbidity (Smith et al, International Journal of Cancer 2019), cumulative burden (Smith et al, Cancer Epidemiology 2020)

Work to examine total morbidity (physical and mental health), cardio-metabolic, and kidney outcomes according to patient characteristic groups, such as age group, sex, ethnic group, calendar period in the Yorkshire region will be carried out and these data will provide the background rate in the general population.

Tabulated mental health data has previously been used been used to make an appropriate comparison of the whole cohort and further help the analysis. This contributed to a completed PHD Thesis. However, going forward record level data will be used.

The following outputs produced are from the combination of data disseminated under this Agreement and NIC-11809-H1Y3W. It is only when using data from both Agreements that the outputs detailed could have been produced.

Summaries of the results will be presented orally at conferences and are intended to be published in academic or medical journals. All outputs will be aggregated with small numbers suppressed and in line with the HES Analysis Guide.

Work describing risks of health effects of treatment in relation to respiratory illnesses was completed by the YSRCCYP research team and published in International Journal of Cancer in 2020. Work describing the applicability of risk-stratified levels of aftercare in predicting long-term morbidity in young adults was completed by the YSRCCYP research team and published in the Journal of Cancer Survivorship in 2020 Work describing risks of health effects of treatment in relation to the cumulative burden of subsequent neoplasms, cardiovascular and respiratory morbidity was completed by the YSRCCYP research team and published in Cancer Epidemiology in 2020.

Further outputs will be disseminated in open-access journals (e.g. BMC Cancer) and presented at conferences including the National Cancer Registration and Analysis Service Cancer Outcomes annual meeting, Teenage Cancer Trust and the International Society of Paediatric Oncology annual conferences. Analyses describing educational outcomes will be submitted for publication by Summer 2023 to the International Journal of Cancer (or similar).

Work will be submitted to the European Journal of Cancer (or similar) in relation to specific cardio-metabolic disease by Summer 2024. Risks of kidney toxicity will be submitted for publication by Summer 2024 in the British Journal of Cancer (or similar) by Summer 2024. Risks of anxiety, depression and overall mental health disorders will be submitted for publication by Winter 2024 in the International Journal of Cancer (or similar).

Summary reports of the work and research undertaken will be compiled and also made available on the Yorkshire Register University of Leeds website (www.ysrccyp.org.uk), according to the timelines listed earlier in the document.

All outputs will be aggregated with small number suppression in line with the HES Analysis Guide.

Further outputs which have made use of the general population data disseminated under this Agreement are as follows:

- The data collected as part of the YSRCCYP has allowed researchers to better understand the significance on NHS specialist services of late cardiovascular and respiratory morbidity as well as second cancers in relation to childhood and young adult cancer survivors and quantifies the long-term risk of such morbidity for survivors and healthcare professionals looking after them.

- The YSRCCYP database has facilitated the production of an up-to-date summary of the latest literature relating to late consequences of cancer treatment on reproductive health, describing the impact on both fertility and pregnancy.

- It has allowed researchers to quantify respiratory morbidities, treatment-related risks and their relationship to subsequent morbidity and mortality among long-term survivors of childhood and young adult cancer in Yorkshire. By age 40, cumulative incidence for an admission for any type of respiratory condition was almost 50%. Respiratory admission rates were almost 2 times higher in cancer survivors than in the general population. Treatment with chemotherapy with known lung toxicity increased the risk of admission for all respiratory conditions. Subsequent mortality was highest in those admitted for pneumonia compared to other respiratory conditions.

- It has allowed researchers to quantify the prevalence and spectrum of mental health problems found in adult survivors of childhood cancer, based on a systematic review of the current evidence. Problems ranged from depression, anxiety, behavioural problems and drug misuse. Factors increasing the likelihood of mental health problems included treatment with high-dose anthracyclines, cranial irradiation, diagnoses of sarcoma or central nervous system tumours and ongoing physical ill health. The review recommended further work to identify childhood cancer patients who are at risk of developing late mental health morbidity.

- The first comprehensive analysis of hospital mental health episodes following CTYA cancer treatment (Friend, PhD thesis, 2020), preceded by a systematic review of the current evidence (Friend et al, Int J Cancer 2018).

Results will be reported to academic audiences in peer reviewed journal articles and conference presentations. The study also plan to disseminate findings directly to other beneficiaries including cancer patients, their families and clinicians.

Data will be held for as long as the research project is funded to undertake this piece of epidemiological and applied health research. Current funding expires on 31st August 2027. Subject to securing ongoing funding, the data would be retained until December 2027 to allow sufficient time for completion of analyses, submission and final publication of papers and as long as data is retained, an agreement will be in place with NHS Digital.

The yielded benefits so far have been yielded using the aggregated HES and Mental Health denominator data which was requested to go alongside linked childhood cancer cohort (disseminated under DARS-NIC-11809-H1Y3W) to draw comparisons with. These were limited to tabulated extracts according to 5-year age groups and sex, and unfortunately it was not possible to undertake proper adjustment for confounders such as ethnicity and deprivation.

Results using the data from both Agreements have nonetheless contributed to a completed PhD thesis by a Clinical Research Fellow at the University of Leeds, entitled “The Long Term Mental Health of Survivors of Childhood and Young Adult Cancers”. A link to a copy can be found here: https://etheses.whiterose.ac.uk/27118/

The University of Leeds requires HES and mental health data for a specific cohort to be analysed, alongside data collected in the Yorkshire Specialist Register of Cancer in Children and Young People (YSRCCYP), to continue its epidemiology and health services research programme.

The justification for processing the data by University of Leeds is Article 6 (1) (E) of the GDPR: (processing is necessary for the performance of a task in the public interest or in the exercise of official authority vested in the controller). The justification for the processing of the special category data (health data) by University of Leeds is Article 9 (2)(J) of the GDPR: (processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes in accordance with Article 89(1) based on Union or Member State law which shall be proportionate to the aim pursued, respect the essence of the right to data protection and provide for suitable and specific measures to safeguard the fundamental rights and the interests of the data subject)

The sole data controller is the University of Leeds. The legal basis for dealing with people’s personal data for research data, is ‘task in the public interest’, or in other words, the University are collecting data that they need in order to complete their research, which aims to improve healthcare for the general public.

For background, the YSRCCYP is a regional population based register containing detailed demographic and clinical information on children and young adults aged 0-29 years diagnosed with cancer since 1974. The YSRCCYP covers the Yorkshire and Humber Strategic Health Authority (SHA) which has a total population of 5 million people. Spanning an area of 15,000 square kilometres the Yorkshire and Humber SHA comprises a range of urban and rural communities with a significant ethnic minority population resident in parts of West Yorkshire.

The YSRCCYP research team, within the University of Leeds, is notified of patients eligible for inclusion in the YSRCCYP either directly by the patient’s treatment centre or via electronic reports from the National Cancer Registration and Analysis Service. The YSRCCYP research team then obtains information on patients by manual data abstractions from hospital records. Detailed data on the patient and diagnosis, including treatment information for each of these cases is obtained by a sole data collection officer via the medical records at relevant hospitals in the area, and annual follow up of all cases takes place to ascertain data on any relapses or deaths through letters sent either to the patient’s treating consultant or general practitioner. Data on 9,500 patients have been collected since 1974, however linked HES and mental health data was required for only 8,500 as the cohort shared with NHS Digital excludes participants who passed away before 1996. The cohort submission under approval of this request will consist of approximately 7,000 patients.

The YSRCCYP was originally set up in collaboration with local clinicians to provide research information. Since 1994, the YSRCCYP database and research programme has been managed by and at the University of Leeds’ Division of Epidemiology & Biostatistics. The University of Leeds is the Data Controller for the YSRCCYP with sole responsibility for determining the purposes for which and the manner in which any personal data are processed. The work is currently funded by the Candlelighters Trust.

The purpose of the YSRCCYP is to facilitate population-based epidemiological and health services research. The use of HES and mental health data contributes to this by providing additional information that can be linked with and analysed with data from the YSRCCYP data. The HES and mental health data are not added into the YSRCCYP research database. The two datasets are stored separately but contain common unique study IDs enabling data to be linked at record level. Where required for specific research, relevant data are extracted from the respective databases, linked and analysed by the YSRCCYP research team.

A current research focus is on hospital burden around the time of diagnosis and treatment and monitoring long term risks of hospitalisation associated with cancer treatment. One specific processing activity will relate to describing the risks and prevalence of mental health illness within the cancer cohort compared to the general population.

This type of epidemiological and health services research has the potential to benefit future patients by identifying risk factors which can be used by health professionals to identify those at greatest risk of mental health illness so that interventions and appropriate support can be implemented. It may also reveal important environmental risk factors, examine changes in incidence rates which may help to identify possible causes and understand survival patterns according to ethnic group and socio-economic status in order to ensure that there are no inequalities in outcomes or access to specialist cancer care for certain sub-populations.

The YSRCCYP research team’s research plans include the following objectives:

1) To describe the total burden of hospitalisation among the Yorkshire cancer population aged 0-29 years, to identify clinical and sociodemographic factors which influence the likelihood of hospitalisation and to investigate how hospitalisation rates have changed since 1997.

2) To understand patient care pathways through the NHS before, during and after cancer diagnosis. This includes assessment of time to diagnosis for children and young adults diagnosed with cancer under the age of 30 years to identify where improvements can be made to minimise delays in diagnosis leading to better prognosis and less stress and anxiety on patients and their families.

3) To calculate the risks and costs to the NHS of adverse health events requiring hospital admission for survivors of cancer in this age group so that clinicians can provide appropriate follow-up care.

To address aim 1 above the YSRCCYP research team will utilise linked HES and mental health data to investigate long term risks of respiratory and mental health illness in the cohort and identify sociodemographic and clinical factors which may affect these risks. The linked HES and mental health data are covered under a separate Data Sharing Agreement (reference: NIC-11809-H1Y3W).The YSRCCYP research team also wish to determine the relative excess risk of these conditions within the cancer cohort compared to the general background population and, in order to make this comparison, requires a separate pseudo-anonymised extract of HES data containing all episodes for patients in the Yorkshire and Humber SHA area only under the age of 60 at admission (the oldest person currently registered). This separate extract is covered under this Data Sharing Agreement.

The YSRCCYP research team require some data items classed as sensitive. These are the Referrer code, which indicates the manner in which the patient was referred to hospital by ascertaining the code of the referring organisation. This allows the YSRCCYP research team to identify particular patient pathways which are associated with an optimal time to diagnosis, a key indicator known to influence survival. Additionally the Consultant code data field is required because it enables the YSRCYYP research team to work out whether patients receive care at specialist cancer centres as opposed to general district hospitals, in order to address important health services research questions such as: ‘Does specialist care improve patient outcomes for children and young people including length of hospital stay and reduce subsequent morbidity and mortality?’. There are currently no databases which link consultant codes to specialist cancer centres for childhood and young adult cancer, so this process needs to be done manually using cohort linked NHS Digital data and the YSRCCYP database.

Update June 2019: For this iteration of the agreement - University of Leeds will receive tabulated extracts of mental health data, including the Mental Health and Learning Disabilities Dataset and the Mental Health Minimum Data Set - covering the period from 2006/07 to 2015/16. University of Leeds will use aggregated data with small numbers supressed, broken down by age and single year (as this allows UofL to determine whether there is an age at which admissions peak or reach a maximum before plateauing), rather than record-level data. Additionally, UofL are only requesting data for the population under the age of 59, as the comparision cohort (from the Yorkshire Specialist Registry of Cancer in Children and Young People) are at most 58 years old now. UofL have not requested data from older people as this will not be used.

UofL are keen to access data by gender, deprivation (quintiles), ethnicity (South Asian, Black, White, other) as well as age to allow them to draw accurate comparisons with the cancer cohort. However, aggregated data will allow UofL to make an appropriate comparison without requiring access to lots of patient-level data.

Work describing risks of health effects of treatment in relation to respiratory illnesses has been completed by the YSRCCYP research team and submitted for publication in the International Journal of Cancer. This work was published in the International Journal of Cancer December 2018 https://doi.org/10.1002/ijc.32066. The December 2018 publication follows a June 2016 publication where descriptive statistics have been produced showing the respiratory conditions diagnosed within the linked cohort. The background admission rates in the general population are required over the same time period to enable further statistical analysis to be carried out. Outcomes of the work will also be disseminated in open-access journals (e.g. BMC Cancer) and presented at conferences including the National Cancer Registration and Analysis Service Cancer Outcomes annual meeting, Teenage Cancer Trust and the International Society of Paediatric Oncology annual conferences. Further work will be submitted to the European Journal of Cancer (or similar) in relation to specific mental health outcomes by December 2019. Analyses describing the variation in clinical pathways including delays and time to diagnosis will be submitted for publication by September 2019 to Journal of Clinical Oncology (or similar).

Additional work describing the rates of hospital activity and differences between ages at diagnosis (e.g. 0-14 vs 15-29) and ethnic group (e.g. south Asian vs non-south Asian) was completed in October 2017 and submitted for publication to the British Journal of Cancer by in December 2017. Details of risk stratification models and the methodology to derive these for individual patients will be disseminated by the research team to every clinician involved in the care of children and young people (CYP) in August 2019. This will be supported by the Yorkshire & Humber CYP cancer network that holds details of all practising NHS CYP cancer teams and clinicians in the region.

Summary reports of the work and research undertaken will be compiled and also made available on the Yorkshire Register University of Leeds website (http://medhealth.leeds.ac.uk/info/545/yorkshire_specialist_cancer_register), according to the timelines listed earlier in the document.

All outputs will be aggregated with small number suppression in line with the HES Analysis Guide.

As the funder, the Candlelighters Trust may request information for use in its own information dissemination and publicity materials. For example, they may ask for the number of new cases diagnosed per year in Yorkshire and projected incidence rates. The University of Leeds would only share information that is available as a result of the processing activities described above – i.e. the YSRCCYP would not undertake further data processing in order to derive information requested by the Candlelighters Trust and any information shared would be put in the public domain. For clarity, the University of Leeds is not obliged to provide information on request to the Candlelighters Trust and would not share any data that are not aggregated with small numbers suppressed in line with the HES Analysis Guide.

The linked NHS Digital data alongside the background hospitalisation rates will be used to derive key information which will be provided by the YSRCCYP research team to clinicians involved in the long-term care of young people identifying each individual’s risk stratification group (defined as being at ‘low’, ‘medium’, or ‘high’ risk of future complications or health effects, based upon their previous hospital activity patterns, treatment mortality, dose, cancer type and stage). The risk stratification model will be devised by the YSRCCYP research team and disseminated to clinicians in the Yorkshire and Humber region via the Y&H Children’s and Young People’s Cancer Network (August 2019). Only those clinicians involved in the direct care of individuals with cancer will be provided with details of the risk stratification model. Oaediatric and adolescent oncology and haematologists will be provided with aggregated cancer intelligence data on the number of survivors currently being seen at each NHS Trust according to risk stratification group, so future services can be planned effectively (June 2020).

Data will be held for as long as the research project is funded to undertake this piece of epidemiological and applied health research. Though work is currently planned until December 2019, the current funding expires on 31st August 2020 to allow sufficient time for completion of analyses.

Several peer-reviewed scientific papers incorporating HES and registry linked data have been published including:

van Laar M, Feltbower RG, Gale CP, Bowen DT, Oliver SE, Glaser A. Cardiovascular sequelae in long term survivors of young people’s cancer – a linked cohort study. Br J Cancer 2014; 110: 1338-1341.

Althumairi A, Feltbower RG, van Laar M, Kinsey SE, Glaser AW, Picton SV. Patterns of hospital admissions and length of stay during 1996 to 2011 among children compared with teenagers and young adults after completing treatment following diagnosis with cancer in Yorkshire. Eur J Cancer Care 2015; 24: pp.9.

Fairley L, Stark DP, Yeomanson D, Kinsey SE, Glaser AW, Picton SV, Evans L, Feltbower RG. Access to Principal Treatment Centres and survival rates for children and young people with cancer in Yorkshire, UK. BMC Cancer 2017; 17: 168

Smith L, Norman P, Kapetanstrataki M, Fleming S, Fraser LK, Parslow RC, Feltbower RG. Comparison of ethnic group classification using naming analysis and routinely collected data: application to cancer incidence trends in children and young people. BMJ Open 2017, 7 (9) e016332; DOI: 10.1136/bmjopen-2017-016332

Presentations of on-going work have been presented to academic and clinical audiences including: analysis of mental health late effects to the Royal College of Paediatrics and Child Health 2018 Conference; analysis of respiratory late effects has been presented at the Public Health England Cancer Services, Data and Outcomes Conference in June 2018.