How much subspecialization do we really need in radiology?

The weary joke about subspecialization in medicine goes that a team providing cutting-edge care for big toe pathology fears obsolescence in the face of increasing big toe technical and academic advancement and eventually splits. Team hallux sinister and team hallux dexter are born.

Like all worn-out jokes, there is an underlying kernel of truth. Healthcare has become more complex. Diagnostic and management pathways have become intricate with multiple decision nodes sometimes requiring high-stakes choices at pace. All this requires staff intimately conversant with the details of the pathways they provide and able to triage to other teams when they recognize a patient falling out of their expertise area: It requires subspecialization.

Dr. Chris Hammond.Dr. Chris Hammond.

Subspecialization has undoubtedly resulted in improvements in the care of patients in much of medicine. If you had a myocardial infarction (MI) in the 1960s, you'd be put on a ward to see if you got better and maybe prescribed some calisthenics if you did. You were lucky if you got aspirin.

Now with subspecialty teams providing 24/7 percutaneous coronary intervention, secondary prevention, risk stratification, cardiac rehabilitation, delayed elective revascularization, and so on, survival and morbidity following an MI are unrecognizable from the 1960s and continue to improve further (see 2017 paper by Saga Johansson et al).

The joke though is that the endpoint of subspecialization is absurdity. There is a risk of disappearing down a technical and professional rabbit hole so deep and long that you and your colleagues become irrelevant – just like teams hallux sinister and dexter. But it's not necessary to reduce to the absurd to highlight a number of risks with subspecialization that temper its undoubted benefits.

The first risk, and the most obvious, is that patients don't come readily parcelled in handy chunks of pathology that fit neatly into clinical pathways. They are complicated. The 80-year-old man with a 6.3 cm abdominal aortic aneurysm may also have prostate cancer, chronic obstructive pulmonary disease, a wife with early dementia, and family that -- while caring and concerned -- lives too far away to offer any robust practical support.

We can all recall patients whose lives become dominated by increasingly frequent visits to hospital, meeting fractured and disparate healthcare teams. How do we prevent their experience of care from becoming confusing and burdensome? Who is going to help them make a holistic decision about what their therapeutic priorities are when each subspecialty team is only familiar with a little bit of their lived experience? As the population ages, complex multimorbidity is commonplace, and patient-centered practice becomes increasingly important. This requires generalists or the active maintenance of a generalist overview. Or perhaps paradoxically, specialist generalists like "POPS" (Proactive care for Older People undergoing Surgery) teams.

A second risk is that it can lead to overtreatment. In the context of the joke, the teams hallux justify their setups and ongoing existence by the need to undertake increasingly complicated interventions on big toes. Need is a slippery word here (see my 2021 blog, What do we mean by need?). Is the need driven by a population health deficit that can be cost-effectively corrected, or by a professional interest in a particular niche pathology? Economists call this supplier-induced demand. In lay terms, if you are a carpenter with a hammer, everything looks like a nail.

Finally, increasing subspecialization can lead to loss of workforce cohesion and siloed – rather than systems – thinking. When colleagues have little in common, they communicate and collaborate less and can even start to perceive each other as threats. Opportunities for informal interactions are lost, relationships deteriorate, and solidarity withers.

This situation can be exacerbated by specialty teams working in geographically isolated locations. But there remain many shared challenges in the provision of healthcare that are common, whatever the subspecialist interest, and which require teamwork and cooperation to solve. These challenges vary in scope and complexity from the contribution by subspecialty teams to out-of-hours or emergency general services to much wider policy or philosophical problems such as population-level health interventions, social justice in health, inclusion and equity in healthcare provision or the sustainability and environmental impact of services.

In my own subspecialty of interventional radiology, conversations about these wider aspects of what we do are very much in their infancy (see March 2023 paper by Anouk de Reeder et al).

In radiology, subspecialization is inevitable and necessary. I can perform transarterial chemoembolization (TACE), but I'm not very good at reporting the liver MRI on which the liver lesion was diagnosed. I think the converse applies. Modern imaging is so complex, so rapidly evolving, and so central to modern healthcare that subspecialization within radiology is a representative microcosm of specialization within medicine in general.

But as in medicine in general, subspecialization in radiology creates difficulties. Even in a large department, comprehensive cover for acute radiology, especially out-of-hours, can be complex to organize if colleagues no longer feel competent to report imaging outside their immediate area of expertise. Cover becomes a complex tessellation of overlapping subspecialty skills, supported by byzantine risk-assessed protocols for who reports what and when, creating confusion for referring clinicians and trainees alike. Or it is outsourced to an external provider willing to provide nonspecialist radiology for a fee.

Subspecialization can also mean that some services, modalities, or examinations get left behind, lost in the gaps between multiple areas of individual expertise. Plain film radiology and general ultrasound spring to mind. Finally, operational pressure in one subspecialist area (for example a particularly large backlog due to planned or unplanned leave or a scanning initiative) is difficult to mitigate if the excess workload cannot be shared more widely.

One of the roles of radiology leadership is to manage these challenges, to steer a department between the competing risks and benefits of subspecialism vs. generalism and of individual aspiration vs operational necessity. The key to this is the nurturing of teamwork and communication.

Subspecialization within teams is something we are familiar with in all aspects of life: in sport, think of the seamless drafting of a cycle team or the elegance and grace of a football team at the top of their game; in commerce, think of the thousands of people with different skill sets it took to design, produce, deliver, and maintain the device you are reading this on; in family life, think of your role as parent, lover, cook, taxi driver, breadwinner, emotional anchor, comedian, or straight man. Subspecialist cooperation is ubiquitous.

Part of the key to a successful radiology team (any team) is not that everyone does the same thing or is treated the same way. It's the creation of a culture of professional respect, equity, and understanding within which individuals or groups can enjoy and deliver their subspecialty interests (with all the clinical and operational benefits this brings) without risking the downsides of this specialization. A radiology team with a strong sense of overarching collaborative endeavor and collective ownership will automatically mitigate the risks of subspecialization, of siloed thinking, of isolationism, or protectionism. It will identify the gaps and manage them.

But a strong team culture is not only built or modeled by its leadership. The members of the team also need to be willing to be team players for it, to sacrifice some (not all) of their personal desire for the sake of the collective. So it's OK to have your aspirations in the clouds, to be an expert left-big-toe radiologist, to have an inch-wide, mile-deep practice some of the time. But you also need to keep yourself grounded in the general, in the shared. You may find that this can be just as fascinating, just as fulfilling. The broadening of perspective it affords as you leave your rabbit hole, see the sky, and breathe the fresh air of new opportunity can be invigorating! I know because I've done it.

So be partially pluripotent. Don't fall into the trap of subspecializing yourself into irrelevance. Avoid this mainly for your own sake, but also for the benefit of your colleagues, your department, and the community it serves and your profession.

Dr. Chris Hammond is a consultant vascular radiologist and clinical director for radiology at Leeds Teaching Hospitals NHS Trust, Leeds, U.K.

The comments and observations expressed herein do not necessarily reflect the opinions of, nor should they be construed as an endorsement or admonishment of any particular vendor, analyst, industry consultant, or consulting group.

Page 1 of 253
Next Page