Data directly from our patients: Is improving patient data the key to better cancer care?
Professor Corinne Faivre-Finn, Professor Niels Peek, and Dr James Price
To provide the best care and support for cancer patients during and after treatment, it is essential to collect and work with data that captures patient experiences and patient-reported outcomes. But data is not a simple subject. The way healthcare services work with data, and how we work with patients to collect it, must be tackled from several angles. So, what can be done and what are the priorities?
Challenges with real-world data
To obtain real-world evidence, for example to establish the outcomes of a particular course of treatment and the lessons we can learn for future treatments, it is essential that researchers can easily access high quality real-world data collected (for example from patient electronic health records) and also patient generated health data. Data should be collected in a structured way both at the start of a patient’s journey (diagnosis) and during their treatment pathway. This will allow the capture of demographic, cancer-specific and survival data that could be comparable to the data collected through clinical trials, particularly if quality control processes are built in. At present, the quality of routinely collected data in healthcare settings is still substandard when compared to that gathered in clinical trials, where strict and rigorous processes provide quality assurance.
One of the challenges faced by researchers is the lack of data collected on the impact of a medical condition, such as cancer, and its treatment from the patient’s perspective. Clinician-reported data has been found to frequently under-report the frequency and severity of symptoms experienced by patients with cancer. A possible solution is to collect patient-reported outcome measures (PROMs) which are self-completed questionnaires on symptoms and quality of life.
Another challenge is the governance associated with real-world health data. Data sharing is more complex since the introduction of the General Data Protection Regulation (GDPR). To overcome this barrier, an umbrella exemption (Section 251 approval), can be applied for, which first requires NHS ethical approval and a sponsor, and then further applications before permission is granted. Obtaining these umbrella approvals via Section 251 to undertake research with large, real-world datasets is challenging and maintaining public trust regarding data sharing and security is paramount.
Finally, while a positive data culture exists, it is currently hampered by a lack of time and training for most clinical teams who will be involved with the collection of the data.
Clinician-reported data has been found to frequently under-report the frequency and severity of symptoms experienced by patients with cancer.
ePROMS: New ways to capture and use data
The routine collection of electronic patient-reported outcome measures (ePROMS) helps to significantly improve the understanding of the side-effects from treatment and the quality of life for all patients.
The Christie Hospital in Manchester has set up ePROMS as a clinical tool to help improve the communication between patients and clinical teams. The aim is to better tailor treatments and aftercare as well as providing high-quality data to facilitate real-world research. This initiative, a collaboration between The Christie, The University of Manchester and Manchester Cancer Research Centre (MCRC), will also help to enhance many other aspects of routine patient care.
Despite the impact of COVID-19 the roll-out of ePROMS to date has been highly successful, and with funding from The Christie Charity, ePROMS will be rolled out to all relevant disease groups at The Christie by the end of 2022. In addition, a real-time responsive service for patients will be developed in response to patients’ reported symptoms. In this way, ePROMS will feed into models to predict clinical outcomes which may help clinicians to make decisions about treatments. ePROM data may also allow clinic appointments to be used more effectively for patients who require clinical input and reduce unnecessary visits for well patients who do not. Patient care pathways are already becoming personalised as a result of ePROM data. For example, patients receiving Herceptin and other anti-HER2 drugs following surgery for breast cancer are monitored with serial echocardiograms and ePROM questionnaires. Patients who report via ePROM that they are well with no symptoms are able to proceed with treatment without clinical review and patients reporting symptoms via ePROM receive clinical review.
Another recent development in improving the accessibility is our UK Computer Aided Theragnostics (ukCAT) project, launched at Manchester Cancer Research Centre and The Christie in 2017. This system makes anonymised routine patient data available to research organisations who can use those insights to develop new approaches to diagnoses and treatments.
Processes for sharing data in a way that are quick, transparent and acceptable to patients and the public is vital to the ambition to progress towards real-world evidence at scale.
Public consultation and how data is shared
Processes for sharing data in a way that are quick, transparent and acceptable to patients and the public is vital to the ambition to progress towards real-world evidence at scale. The University of Manchester has been working alongside national bodies, including the National Institute for Health Research (NIHR), NHSX, and NDG (National Data Guardian) to run public consultations on the advantages and risks of data sharing.
A recent public consultation, focused on health data sharing in the pandemic, used three citizens’ juries to weigh the benefits of valuable data systems against disadvantages of continuing to process data. Jury members were informed about what data sharing and electronic health records are, why they are important and what the data is used for. This way of bringing complex evidence to the public enabled juries to make judgements and reach reasoned answers to questions concerning data.
The consultation revealed that overall, juries supported decisions to introduce data initiatives during the pandemic and the majority were in favour of all data sharing initiatives continuing, as long as they were valuable. In addition, 94% of jurors believed that an independent body of experts should review data sharing initiatives, rather than those organisations responsible for the initiatives.
Additionally, MCRC’s Patient and Research Data Statement exemplifies the commitment to an ambition of using patient data in a safe and secure way to look at new innovative ways to improve patient outcomes. This has been accomplished through best practice related to patients in research programmes with strict security measures that protect patient confidentiality. Data collected includes real-world data combined with the Greater Manchester Care Record to improve individual cancer care, including diagnosis, treatment and monitoring of cancer. Our system uploads the symptoms, quality of life and experience of patients directly to central and secure data storage at The Christie. This can then be used by both clinicians and patients for them to jointly make informed clinical decisions.
MCRC's Patient and Research Data statement exemplifies the commitment to an ambition of using patient data in a safe and secure way to look at new innovative ways to improve patient outcomes.
Patient data and steps for the future
There are very few places that have been able to establish ePROMS, or similar patient data initiatives, within a routine setting. Such initiatives require high levels of expertise, engagement with patients and staff, validation of work and in-depth data analysis. To make them a successful and integral part of our own health and care system, we need to take a number of preliminary steps:
- Patient data must be properly linked between primary and secondary care. This results faster communication between both parties and ensures that all relevant information is available in both settings. Leadership from Department for Health and Social Care, NHS England and from our newly developing Integrated Care Systems is key to building this ability.
- To help develop a robust research culture, both current and future clinicians will need further training to understand the importance and clinical relevance of high quality data collection. Organisations like NHSX and Health Education England can play a leading role in setting these standards and delivering the necessary workforce and student training. This should involve a review of current provision of both data collection and quality assurance training, in addition to consultation with researchers and clinicians to develop new content appropriate to our rapidly advancing technologies.
- The process for obtaining permission for the use of data in clinical research must be made easier: opportunities are currently being lost because of the time and complexity required to access the data that researchers need to develop new techniques and treatments, and to co-ordinate healthcare more effectively between organisations within the NHS. We need to explore whether there is space here for exemptions from GDPR and for other regulations to be centrally administered, rather than on a piecemeal basis (where each NHS Trust provides individual exemptions for its data). This must, of course, be accompanied by safeguards and oversight of the highest standards, to ensure public confidence, but these safeguarding standards are already being developed in universities and NHS Trusts across the country and will not be a case of reinventing the wheel.
- We should also listen to our citizens’ juries, who have suggested the formation of an independent body of experts to oversee health data sharing within the NHS and with the research community. This would require a collective initiative, led by central government, to build and establish, but would be supported not only by research organisations around the country but by informed public opinion, as evidenced by our juries’ conclusions.
- It is also essential that data collected from electronic health records can be linked with other digital sources. This requires pragmatic adaptations to GDPR and other data protection laws and working with citizens’ juries to understand the public’s response to new forms of data sharing in the health and social care system.
As we look to the future, building and maintaining public confidence in the integrity and effectiveness in the use of patient data must be understood as the first and essential condition for their future development. Without this, we cannot fully benefit from the promise of real-world evidence research and the great advances in research and treatment that they hold.
About the authors
Professor Corinne Faivre-Finn
Corinne Faivre-Finn is Professor of Thoracic Radiation Oncology, The University of Manchester and Honorary Consultant Clinical Oncologist at The Christie. Her interests lie in the development of advanced radiotherapy techniques, combined modality treatments for non small-cell and small cell lung cancer, real word and electronic patients-reported outcomes.
Professor Niels Peek
Niels Peek is Professor of Health Informatics in the Division of Informatics, Imaging and Data Science (School of Health Sciences), and lead for Digital Health and Care at the Christabel Pankhurst Institute for Health Technology Research and Innovation. His research focuses on translational data science for risk prediction, personalised and precision medicine, patient safety, and multimorbidity.
Dr James Price
Dr James Price is a PhD Clinical Research Fellow at The University of Manchester and Consultant Clinical Oncologist at The Christie, where he specialises in the treatment of patients with cancers of the head and neck. He is the Clinical Lead for ePROMs roll-out at The Christie.
Rules of the road: The need for new quality standards for AI technology in healthcare
Dr Dónal Landers, Dr Gareth Price