Scanners, shortfalls and stark statistics

Investment in diagnostic scanners: will it shift the dial? 

After years of under-investment, recent government announcements on capital spending in the NHS make a welcome change. In the autumn budget, Labour seemingly reconfirmed their manifesto pledge to double the number of CT and MRI scanners by revealing £1.5 billion of funding for these (as well as for surgical hubs and hospital beds). This is positive news and will address a real need in the system.

Most of the budget’s healthcare content was explicitly linked to the aim of bringing down NHS waiting lists. However, if the government thinks that increasing our complement of scanners will suffice by itself to do this, then they are in danger of disappointment. The truth is that there are multiple bottlenecks – each created by a stark mismatch between demand and capacity – at each stage of a patient’s pathway, from investigation to diagnosis to treatment to discharge. While it is true that one bottleneck is getting a patient scanned in the first place, a second, and arguably more severe, bottleneck is in the analysis of the images produced by those scans to obtain a diagnosis. This second blockage is created by the now almost endemic shortage of clinical radiologists across the country. 

We know that we have, right now, a 30% shortfall of clinical radiologists in the UK. That means we have 30% fewer radiologists than are required to meet the demand for reporting. Moreover, the country is facing a precipitous further decline in the workforce, as we predict the shortfall to rise to 40% by 2028 (if no action is taken). Quite simply, diagnostic waiting lists will continue to remain sky-high until there are enough radiologists to clear the backlog.

If the government’s foremost healthcare aim is to restore performance against waiting times targets, then they need to recognise that scanners are necessary but not sufficient for the task. Rather, it is clinical radiologists and diagnostic scanners together that are jointly necessary and sufficient. We need to labour this point because it is not always clear that those in government understand what a radiologist does and the centrality of their role. Comments made in the House of Commons suggest that there is an assumption that investment in scanners alone will solve the problem.

Indeed, there is a risk that by increasing the number of scanners the blockage at the reporting stage is exacerbated, rather than ameliorated. If more scans are produced each day, without commensurately expanding reporting capacity, then radiologists’ workload is increased. Radiologists worry that in this scenario, to maintain or improve performance against targets, they will be expected to work even harder – at a time when many are reaching their limit. 

This risk is particularly clear if these new scanners come with the latest inbuilt artificial intelligence (AI) software. The last government revealed NHS figures estimating that AI-upgraded MRI scanners would each produce an additional 3.71 scans per day, by virtue of their increased scanning speed. The RCR analysed these figures and revealed that 100 such scanners would therefore require an additional 32,500 hours to report – equal to the work done by 40.6 full-time consultant radiologists.

Clearly then, without a commensurate increase in reporting capacity, any increase in image acquisition capacity will at best have little effect on waiting lists – and at worst, actively make them worse by piling the pressure on the reporting stage and on our already over-extended radiologists.

Of course, one solution will be to introduce AI tools at the reporting stage too, with the aim of speeding up the time taken for a radiologist to analyse and report each scan. The promise certainly exists for using AI software for this purpose, and the RCR has set out clear actions the government should take if it wants to increase its uptake. However, even in an ideal scenario in which this AI to augment diagnostic imaging is introduced effectively and works exceptionally well – and that scenario is by no means a foregone conclusion – there would remain the need to increase the radiology workforce. 

The reason for this is simple. Contrary to common belief, radiologists do far more than sit in darkened rooms and report scans. Indeed, there is far more to reporting than a simple yes/no decision concerning the presence or absence of disease or injury. Not only are radiologists increasingly doing more therapeutic work. They also bring their rich experience and knowledge to their reporting work, going beyond the results of a scan itself and augmenting their reports with this information. They are the glue that binds all the specialties together, communicating with them, interpreting ever-more complex data, and putting the findings in the context of a real person who may have many complex health conditions. 

So even if AI at the reporting stage boosts radiologists’ efficiency and speeds up their work, radiologists themselves remain crucial. Indeed, it is likely that the greatest productivity benefits that will be unlocked by AI will be in the administrative sphere – ambient transcription of notes, automatic rostering, generation of letters, and so on. 

The government and NHSE therefore need to set out clearly how they intend to increase recruitment into the radiology workforce and improve retention, in line with this new policy to double the number of CT and MRI scanners. Our annual census reports would be an excellent place for them to start.

The budget filled in a lot of blanks surrounding the government’s agenda for health. But there is much else about the government’s plans that could and should be discussed. For instance, how much of the £1.5 billion will be spent on diagnostic scanners? Will new scanners be preferentially provided to regions that currently have too few, or whose existing scanners are ageing? How will funding be distributed and how will bids be approved? It is our hope that the Department of Health and Social Care will set out further details of their plans as soon as they are able.

Investment in scanners, and in the NHS estate more widely, is a good thing. Many groups have called for such investment for years and are grateful that the government has listened. Nonetheless, having more scanners does not equal lower waiting lists. The recipe for success on that front is instead: 

X(more scanners) + 3X(more radiologists) = lower waiting lists.   

(Pick your own ratio of scanners:radiologists here. The crucial point is that the radiologist element of this side of the equation is the weightier.)

AI has a role to play in augmenting both the image acquisition and the reporting stages but will not obviate the need to expand the radiology workforce. 

With 330 MPs elected for the first time in 2024, an important task for the RCR will be to increase the name recognition of radiologists and oncologists amongst this group and foster a greater understanding of what our Fellows and members do. This will include dispelling the notion that workforce shortages can be alleviated by greater automation of healthcare delivery.

We will also endeavour to get answers to the questions posed here and to work with the government to bring down waiting lists, increase the number of early diagnoses, move care into communities, and create better outcomes for patients.