• Categories

blog

5 January 2021

Using real-time data and predictive analytics to improve cancer outcomes

Andi Orlowski, Director of the Health Economics Unit; James Peach, Commercial Director of DATA-CAN: The Health Data Research Hub for Cancer; and Deborah Lancaster, Head of Market Access Oncology at Novartis Pharmaceuticals UK, recently joined forces to present a data-focused session as part of an HSJ Virtual Cancer Forum.

The aim of the HSJ forum was to examine the impact of COVID-19 on national cancer care plans and explore how to work together to get back on track. The session on data analytics outlined how it can support a real-time response to a crisis as well as longer term cancer care planning.

James Peach’s presentation outlined the work of DATA-CAN (www.data-can.org.uk) in developing high-quality, connected health data that is accessible for cancer researchers, clinicians and other health professionals across the UK. DATA-CAN also works with patients, the public and health professionals to ensure that data is used transparently and responsibly.

A rare positive impact of the Coronavirus pandemic

James said: “The arrival of COVID-19 had a terrible impact on cancer services, diagnosis and treatment. However, one of the rare positive impacts of the pandemic was that health organisations were more willing to share their data to support the national response to the crisis. When we reached out, eight hospitals from across the UK agreed to share real-time data on their urgent referrals and chemotherapy activity. We were able to then share that collated data openly within a month.”

In addition to analysing activity data, DATA-CAN has worked with partners, including the UCL Institute of Health Informatics, to model the impact of the pandemic and identify the worst affected cancers. The data allows healthcare providers to monitor which kinds of patients aren’t being referred to hospital and therefore being missed.

“We saw for example that the difficulty of developing a COVID-secure scoping practice led to colonoscopy rates dropping to eight per cent of pre-COVID levels in April,” added James. “This recovered to 60 per cent in July. CT scans also dropped to 55 per cent in April, returning to 95 per cent in July.

“Knowing what kinds of cancer may have been missed during the lockdown period is crucial in delivering the right diagnostic services now and catching up with those diagnoses as soon as possible, for the best possible outcomes.”

James says that NHS trusts have increased access to COVID-friendly radiotherapy plans and drugs that require fewer hospital visits with immuno-suppressant side-effects.

“Data is only important if it represents reality and leads to actions that improve outcomes,” added James. “It’s also crucial that we never do anything without engaging with patients around data use.”

Planning for future cancer care

Andi Orlowski’s part of the sessions focused on how local and regional linked datasets, which give a holistic view of people’s lives across health and social care, can add insight to national analytics.

He said: “We can use a predictive algorithm to show how many people are at risk of developing cancer based on the number of patient characteristics. This allows us to target health interventions directly at those at risk of becoming ill.

“We are still seeing health inequalities across the country, as people in our most deprived communities are more likely to die of cancer and to cost more with an over-medicalised death. By targeting those people most at risk of developing cancer using a risk score based on factors such as a family history, weight loss, smoking and alcohol intake, we can begin to address these inequalities.”

Andi warns that the data itself may be biased because the less deprived, wealthier populations have better access to healthcare, so show up more in the treatment data. “In addition to the possible bias, it’s also crucial that we identify those people who are amenable to change, to make a real difference.”

Using data to address unmet needs

Deborah Lancaster said that Novartis sponsored and joined this session as it recognises the importance of data in understanding the patient population, developing treatment programmes and addressing unmet needs to improve cancer outcomes in the UK.

She said: “Looking at how clinicians and patients are currently managing treatment is invaluable to improving the lot of cancer patients. By analysing data that is anonymised at population level we can see how to improve our offering for patients.

“The key is that we can use such data to improve treatment pathways, medicine delivery, patient experience and outcomes. There is a lack of standardisation across cancer centres, and if we can all learn through data insights we can work to iron out the geographic inequalities.

“This HSJ event was an important opportunity for a range of healthcare professionals. Being delivered online, it was compressed and relevant, meaning people could afford the time to access this valuable resource.”

Moving forwards

James and Andi work closely on projects using their respective expertise in data sets and analytics. If you would like to join in their work, please email andi.orlowski@nhs.net or James.Peach@uclpartners.com.

Register to join the second HSJ National Cancer Forum event (www.cancerforum.hsj.co.uk/) on 11 March 2021 which will include input from the Health Economics Unit, James and Deborah.

Our specialist services

This is a small selection of all the solutions we can provide.

Evidence generation

Understanding whether new care pathways and interventions are effective, efficient, and deliver value for money

Population health management

Using allocative efficiency techniques and population health analytics to improve value and deliver the best care possible

Advanced analytics

Using advanced techniques in machine learning, data science and casual inference to understand the biggest questions in health

Consultancy

Sharing our vast knowledge to develop NHS capability through training, research design advice and quality assurance