Monica Jones CITP FBCS, Chief Data Officer at the University of Leeds, shares her experience of improving cancer outcomes through timely data analysis – following the COVID–19-influenced decline in early-stage cancer detection and diagnosis.
When the global pandemic hit, it brought into sharp focus the need for timely intelligence to inform urgent decision–making, when dealing with a virus that spreads rapidly through a population and places huge demands on health services. The UK has a highly regarded public health system, particularly in cancer, with long established disease registries and health data reporting.
However, it became apparent that the system was unable to determine direct and indirect impacts of the virus on cancer patients in a timeframe to underpin rapid action. This was primarily due to the lag time between health data collection, curation and information being made available.
What datasets were used to support your hypothesis?
Health Data Research UK (HDRUK) is the national institute for health data science. At the time I was the Chief Data Officer for DATA–CAN, the HDRUK hub for cancer. We identified this gap and provided real–time data from NHS trusts to demonstrate to physicians, policy makers and governments the significant impact that COVID–19 was having on cancer services UK–wide. This case study highlights how real–time data was deployed to provide crucial intelligence that informed action and underpinned an aspiration to establish a Real–time Data Network (RTDN), which catalysed the use of data to inform better outcomes for cancer patients.
Did you find the correlation by accident?
During the first lockdown (March–May 2020), DATA–CAN and UCL’s Institute for Health Informatics worked with UK cancer centres to collect and analyse real–time hospital cancer service data. During the COVID–19 real–time data research project, we looked at measures to delineate the pandemic’s impact – ‘two week wait’ (2WW), which is referred to as a ‘red flag’, meaning that rapid action is required to save a life. These are referrals from GPs for people with suspected cancers (a reliable measure of the pandemic’s impact on diagnostic pathways), and chemotherapy appointments (a reliable measure of impact on therapeutic pathways). Bringing this data together improved its quality, creating new aggregated datasets that indicated, for the first time in the UK, the impact of COVID–19 on cancer services.
Tell us about the DATA curation and analysis
Curation and analysis of the data found that, compared to pre–COVID–19 levels, there were significant drops in urgent referrals (70 percent decrease) and chemotherapy treatment attendances (40 percent decrease). We employed these results to inform different modelling scenarios of the impact of COVID–19 on excess death in people with cancer, in a cohort of 3.8 million patients with linked primary and secondary care data (the CALIBER dataset).
Our expertise in curating and analysing this data on 24 cancers and 15 co–morbidities enhanced the quality of this dataset and permitted, for the first time, prediction of the magnitude of the pandemic’s impact on excess cancer deaths. Our analysis suggested that the COVID–19 pandemic could result in 7,165 more deaths in people with newly diagnosed cancer. This could rise to 17,910 additional deaths if all people currently living with cancer were considered, with 78 percent of excess deaths occurring in patients with >1 comorbidity.
From a data improvement perspective, our expertise and research activities helped to create new aggregated datasets that did not exist before (2WW and chemotherapy attendance datasets), enhance existing datasets (CALIBER dataset) and deploy these improved datasets to precisely delineate for the first time the impact of COVID–19 on cancer services and cancer patients in the UK.
How did the study become a national news story?
Our rapidly assembled pre–print was shared with the UK’s four chief medical officers, the national clinical director for cancer for England and Scientific Advisory Group Emergencies, contributing to decisions to restart cancer services. The pre–print received more than 22,000 reads with over 80 citations (Google Scholar). The full research paper is available in BMJ Open. Significant media coverage resulted – more than 400 stories in print/broadcast media (including BBC Panorama/BBC Spotlight programmes, lead story for Sky News, Euronews, CBS News; front page articles (Times, Daily Telegraph), with real impact on public opinion.
Be part of something bigger, join BCS, The Chartered Institute for IT.
DATA–CAN findings were quoted in the Health and Social Care Select Committee report ‘Delivering core NHS and care services during the pandemic and beyond’ (October 2020) and during a Westminster Hall debate. Internationally, study data presented to the board of the European Cancer Organisation (ECO) and to the 2020 session of World Health Organization Europe. DATA–CAN’s scientific lead, Professor Mark Lawler, was invited to co–lead ECO’s Special Network on COVID–19 and cancer. This Network launched a ‘7–Point Plan to Build Back Better’, which received significant international attention. The data was also presented to the European Beating Cancer Committee in the European Parliament and will front a pan–European COVID–19/cancer campaign.
Our research was recognised by Health Data Research UK as ‘Highly Commended’ for the Impact of the Year 2020 award
We presented data to the Clinical Research Coalition, chaired by Baroness Blackwood, which made a series of recommendations to Life Sciences Minister Lord Bethell and resulted in this white paper, which highlighted how real–world evidence (RWE) can be used for the benefit of patients, recognising that initiatives like DATA–CAN can help the UK to leverage RWE, become a world leader and set best practice globally. This project, the enhancement of the datasets achieved and its national and international impact, has highlighted to DATA–CAN the absolute primacy of near real–time data. It is also helping to engender a culture of enhanced data sharing in NHS trusts.
Our work has been impactful in multiple sectors. For the NHS, it highlighted for the first time the impact of COVID–19 on cancer services and demonstrated how different trusts could work together to aggregate data for clinical and patient–focused insight. In industry, it has gained significant support, with the Association of British Pharmaceutical Industries highlighting it as an example of how UK datasets and data research can underpin relevant insights to industry.
DATA–CAN researchers have delivered keynote addresses at industry–focused fora and contributed significantly to white papers and reports. From a policy perspective, the work has helped to underpin restoration of cancer services. From a patient and public perspective, it was the lightning rod that first highlighted the negative impact of COVID–19 on citizens and patients. From an international perspective, it underpins ECO’s European COVID–19 campaign with impacts at European Parliament level.
I am proud of my work during the pandemic and I am confident that data did indeed save lives.
About the author
Monica Jones, Chief Data Officer, University of Leeds, and Associate Director for HDR UK, is a former British Army officer with a longstanding interest in technology and science. She moved into health data science after realising her skills and experience could make a difference to clinical outcomes. She has always been interested in science, wanting to know about the natural world and our place in it.