Planning the Radiology Workforce for Cancer Diagnostics

Abstract

The publication of the Delivery plan for tackling the COVID-10 backlog of elective care (NHSE/I, 2022:5) contained a number of ambitions, including that, by March 2024, 75% of patients who have been urgently referred by their GP for suspected cancer are diagnosed or have had cancer ruled out within 28 days. By March 2025, waits of longer than a year for elective care should be eliminated and 95% of patients needing a diagnostic test should receive it within six weeks. The report acknowledged the need to grow the workforce to achieve these ambitions and ensure a timely cancer diagnosis, while also proposing the use of digital technology and data systems to free up capacity. To assist West Yorkshire National Health Service (NHS) organisations to meet these ambitions, this report presents the findings of a ‘deep dive’ that focuses on the role of radiology in meeting the ambitions of providing timely cancer diagnosis. Aims 1. To understand current and projected demand for radiology expertise in cancer diagnosis in West Yorkshire. 2. To understand the current and projected radiology workforce in West Yorkshire and determine the gap between the projected radiology workforce and the required radiology workforce. 3. To identify possible solutions to assist in providing the radiology workforce required for West Yorkshire and explore their acceptability and potential impact. Methods A range of sources of data and methods were utilised. We examined publicly available quantitative data concerning cancer waiting times and diagnostic waiting times and activity and used this to forecast future cancer waiting times and diagnostic waiting times and activity. We examined data from Health Education England (HEE) regarding radiologists’ and radiographers’ workforce profile data for West Yorkshire, the number of radiologists completing training, and the number of radiographers graduating, and data submitted by West Yorkshire Trusts to HEE regarding their plans for growing their radiology and radiographer workforce. Interviews (N=15) conducted with radiology service managers, university academics and key strategic and operational stakeholders delivering radiology services were used to understand the current and future issues around strategic workforce planning, workforce changes and transformation, workforce roles and skills, training and education and service changes. A rapid review of the literature examining the impacts of artificial intelligence (AI) on the workload of radiology services was also undertaken. To put this work in context, we also reviewed relevant policy documents and reports. Alongside this, we consulted with the Yorkshire Imaging Collaborative (YIC) and the West Yorkshire Cancer Alliance (WYCA) and attended a series of workshops run by the Yorkshire Imaging Collaborative. Results Overall, the findings show that demand for radiology services is increasing and that both cancer waiting times and the waiting times for diagnostic tests increased, with a concurrent downward trend in activity that, if all else stays the same, is forecast to continue up to 2025. The cancer waiting times data indicate that patients were waiting longer and that their needs were not being met. Moreover, 3 the proportion of people treated within accepted cancer waiting times decreased both nationally and within the West Yorkshire region from 2013. This was exacerbated by COVID-19 which caused a further decrease nationally and for the West Yorkshire region. National data for waiting times for all diagnostic tests show a significant decline between 2006 and 2008, with a decrease in median waiting times from just under 6.0 weeks to approximately 2.0 weeks. Overall, waiting times remained stable until late 2020 when they started to rise with the longest median waiting times at just over 8.0 weeks in mid-2020. The total number of people waiting for radiology tests nationally is decreasing and is predicted to continue to do so, while in West Yorkshire the number of people waiting for radiology tests decreased until 2020 but has since been on an upward trend which is predicted to continue. Nationally, the total number of radiology tests is on an upward trend that is predicted to continue, while in West Yorkshire activity has been decreasing since well before COVID-19 and is predicted to continue to do so. Data examining the current and future workforce showed that the national figures for the total radiology and radiography workforce are small relative to other health professional groups. In West Yorkshire, 265 radiologists and 926 radiographers were employed, and staff turnover was generally low. Trusts’ forecasts for the number of radiologists and radiographers they believe they need suggest a 16% increase in the number of radiologists in post between March 2022 and March 2027 and a 25% increase in the number of radiographers in post. The numbers of radiographers and radiologists being trained in West Yorkshire suggest that this is feasible. Interview data identified a number of main themes and associated issues: delivering diagnostic cancer targets, strategic workforce planning, workforce roles and skills, service transformation, recruitment and retention, universities, artificial intelligence, collaboration, and international recruitment. Across all themes, some reoccurring issues were identified: a lack of staff, increased demands, a lack of capacity in terms of space and staff, a lack of strategic workforce planning with a focus on operational or financial plans. Respondents proposed potential solutions to some of the issues raised that included: new ways of working, upskilling, developing current and emerging roles, Community Diagnostic Centres (CDCs), greater collaboration between NHS Trusts, universities, CDCs, imaging academies and networks and the private sector, and the international recruitment of radiologists and radiographers to address workforce gaps. The rapid review findings helped to identify a number of potential benefits of use of AI in radiology, including contributing to improved workflow efficacy and efficiency of radiology services. However, this is dependent on the nature of the work and the AI function. As a result of faster AI reading, radiologists may be able to focus more on high-risk, complex reading tasks. AI can support automation of image segmentation and classification and aid the diagnostic confidence of less experienced radiologists. Respondents’ views on AI were mixed. There was acknowledgement that AI was already used to support radiology service delivery and both the benefits and problems associated were identified. The implications of AI for radiologists’ and radiographers’ roles were discussed in terms of changing work, AI being used to support or in some cases substitute radiologists and radiographers, and the need for the radiology workforce to adapt to the technological change whilst maintaining a caring service