Data can help us transform cancer care, but are we there yet?

How are data being used to improve patient outcomes?

Collecting and using data in the most efficient way can help identify deficits in care pathways. In England, where survival rates are worse than the European average in great part due to late presentation, data are being used to improve the speed of cancer diagnosis. The Routes to Diagnosis Programme linked multiple sources of data to examine demographic, organisational, service and personal reasons for delayed diagnosis. Findings revealed that far too many people were first diagnosed in an emergency department, when their cancer was already more advanced. Lessons from the programme were applied to improve cancer diagnosis in primary and secondary care.

One of the most exciting, but largely untapped, areas of data is drawing insights from multiple sources through data analytics, such as artificial intelligence (AI). The power of AI applications can be transformative, but we have a lot to learn before we can fully integrate them into mainstream healthcare. We need adequate analytical methodologies, validated AI algorithms, and ways to avoid biases inherent in data analysis. Addressing these challenges is, however, certainly within our grasp.

One of my favourite examples is the use of AI for image recognition. AI can process images of skin lesions uploaded to a smartphone. It can distinguish between harmful melanoma and benign moles with accuracy similar to that of a dermatologist. Combining human and AI capabilities to classify skin cancers may offer a superior, and more resource-efficient, option for large-scale skin cancer screening.

For those of us who have been advocating for greater use of quality-of-life and patient-relevant data to guide clinical decisions, the launch of the Patient Reported Indicators Surveys (PaRIS) initiative by the Organisation for Economic Co-operation and Development (OECD) is incredibly exciting. The project aims to make healthcare systems more patient-centred through collection of internationally comparable data on what matters most to patients.

Looking towards a future of optimal and equitable data use

The potential for data to ensure greater continuity of care, providing the glue between often disparate providers, is phenomenal. Considerable knowledge can be gained from combining different data sources, including epidemiological, genetic, biological, clinical and socioeconomic data. And the depth of insights we could draw from systematic observation of patients throughout all stages of their care could be transformative, allowing us to achieve truly person-centred care, in real time, for everyone.

Contemplating a world where we make the most of data is an exciting prospect, but it comes with its frustrations, as we look at the systemic barriers that prevent this from materialising. As governments around the world scramble to ‘build back better’ in the wake of the COVID-19 pandemic, we have a real opportunity to take a deep breath and try to address these barriers in a systematic way.

One of the conclusions of our report is that data should be seen as an innovation on equal footing to advances in diagnosis or treatment. Equity of access to data must be a core concern as we move to data-driven healthcare. We need to move from pockets of innovation and build health systems where appropriate collection and use of data are the norm in all settings, and for all patients. Like any major transformation in healthcare, finding usable solutions will require all stakeholders to work together – with patients firmly at the centre.

Find out more about our work on Harnessing data for better cancer care.

The opinions expressed in this blog are those of the author and do not necessarily represent the views of The Health Policy Partnership.