Developing the cancer-screening services of the future
2 June 2019
With cancer incidence increasing as a result of an ageing population, debate as to how the health system could manage this disease burden is rising up the agenda. A focal point for discussion is diagnosis initiatives to identify people with cancer earlier and improve chances of longer-term cancer survival.
A new ‘faster diagnosis standard’, designed to diagnose patients within 28 days will be introduced in 2020, and in October 2018 the Prime Minister announced an ambition that 75% of all cancers would be diagnosed early (stage I or II) by 2028. Screening will also have an important role to play in these efforts: an independent review of cancer screening programmes in England, led by Prof Sir Mike Richards, is currently undertaking an assessment of current screening programmes and the changes needed to future-proof these services.
The Richards independent review has now published an interim report of emerging findings in which he applauds the dedication of the workforce and the number of lives saved through screening. PHG Foundation submitted written evidence to the review in which we focused on the role that emerging technologies could play in supporting screening programmes and the impact that this might have on the workforce.
Potential for risk-based screening
The report highlights the fast pace of research that can support and inform screening programmes, particularly for cancer risk assessment. Developments in tools that incorporate genetic risk scores could contribute to improving screening programmes by pinpointing people most likely to benefit from screening. Our evidence was based on work we have been doing for the EU funded B-CAST (Breast CAncer STratification) project, in which we are exploring the future of risk-based screening in detail. Our reports examining the current policy landscape and detailing the discussions at an expert workshop examining the future of breast cancer screening have already been published. We will be making recommendations based on our analysis of this field in 2020.
Artificial intelligence, and its impact on healthcare, is high on the health system agenda and could be used to develop more accurate models for estimating risk. The speed of development of such AI applications provides a fast moving target that is likely to challenge the slower moving world of regulatory change – challenges which we outlined in our Personalised Medicine Technology Landscape report. The Richards’ review highlights ongoing research into the use of AI to analyse mammography screens. Implemented effectively it could increase capacity of the breast cancer screening service. However, as with genetic risk scores, AI-based tools must be developed with data that captures the diversity of populations.
The interim report highlights challenges faced by the workforce tasked with delivering screening programmes. For example, the roll-out of a new bowel cancer screening programme has been delayed by colonoscopy capacity not increasing to meet demand; pathology services are facing similar pressures. In our written evidence, we argue that while the new National Genomic Medicine Service will boost the genomic analysis of cancer patients and identification of people with inherited cancer syndromes, this will in turn increase the burden on an already stretched workforce. Health professionals urgently need training in managing genomic information and genomic literacy needs to be considered in the context of wider workforce planning if screening programmes are to be delivered effectively.
How could screening look in 2028?
The interim report touches on access to research, data and new technologies to support screening programmes. In our evidence, we put greater emphasis on the potential to harness big data to improve understanding of the epidemiology and pathophysiology of different cancers, to better predict risk, and to develop more accurate screening tools. We also envisage that patients may wish to access services from more diverse sources, such as private providers, and future screening systems could consider mixed models of screening delivery. Delivering this type of service will require investment in digital infrastructure to capture, analyse and integrate data. Collaboration between NHS and non-NHS providers will be necessary and considerable work will be needed to deliver interoperability both within the health system and with external stakeholders.
Whilst confirming the interim status of the findings so far, Prof Richards makes two clear recommendations – that the decline in the number of women attending breast and cervical cancer screening is halted and that action must be taken at a national level to ensure IT systems are fit for purpose. The final report is due later this year and will delve deeper into these and other challenges. What is clear at this stage is that there is an opportunity to support and refine screening services, optimising use of new technologies to support the health system in achieving its early diagnosis goals.