Healthcare Doesn’t Just Have a Radiologist Shortage. It Has an Expertise Allocation Problem
At DICO by Expert Radiotech, we often consider scenarios like this: an MSK radiologist encounters patients imaged at outside facilities whose reports face multi-day delays. The local specialist could interpret the study, but rigid site boundaries and isolated worklists made that expertise inaccessible. While expanding the talent pipeline takes years, healthcare systems can relieve capacity constraints by routing existing expertise more dynamically.
Healthcare leaders should recognize that headcount is not the same as diagnostic capacity. Two healthcare systems with the same 100 radiologists can have dramatically different diagnostic capacity. If subspecialists spend time on cases that do not require their expertise while complex studies wait elsewhere, headcount remains unchanged but effective capacity falls.
Uneven distribution amplifies allocation losses
Radiology needs more trained personnel. The Royal College of Radiologists’ 2025 census reports a 32% clinical radiologist shortfall across the UK, with department leaders linking these gaps directly to delayed diagnoses. Crucially, the national shortage lands unevenly. Detailed RCR policy data reveals that small acute hospitals face a 43% shortfall compared to 30% at large teaching hospitals, while regional gaps in England range from 22% in London to 43% in the North East.
National headcounts fail to reveal whether a department’s immediate constraint is absolute supply, a missing subspecialty, shift alignment, or a worklist isolated from external capacity.
The worklist decides where expertise goes
Uneven workforce distribution becomes more damaging because reading queues are still designed around organizational boundaries.
Every worklist makes resource-allocation decisions. The IHE radiology framework defines worklists as systems where reading clinicians are the primary resource, balancing key criteria such as clinical urgency, reader workload, physician availability, and subspecialty expertise. However, these allocation rules frequently halt at organizational boundaries. Studies remain tethered to a specific site, PACS, or group, creating scenarios where a qualified specialist handles routine cases in one queue while a complex study needing that exact expertise sits stalled in another.
As practitioner discussions on reading outside subspecialties highlight, teams compensate through informal human networks (“phoning a friend” or funneling cases manually) – a process reliant entirely on personal relationships and unmanaged free time.
Allocation gaps accumulate on individual radiologists
Because worklists fail to route cases across organizational boundaries, the structural friction of isolated queues falls directly onto individual physicians.
Fragmented design degrades into human fatigue. A single radiologist often context-switches across subspecialties, fields urgent calls, retrieves priors, and provides informal second opinions while maintaining primary worklist accountability.
While broad general practice is essential (as detailed by Howard Fleishon, MD, in an ACR podcast on community division practices), relying on unthrottled interruptions introduces friction. Unmanaged central worklists risk bottlenecking specialists unless the system distinguishes broad case ownership from narrow, targeted expert input.
Better assignment can recover effective capacity
Reconfiguring this allocation model cannot manufacture new physicians, but intelligent routing directly relieves staff pressure by recovering capacity lost to misassignment and constant interruption.
The operational impact of automated assignment is concrete: a 2024 study looking at the automation of CT assignments at a tertiary cancer center demonstrated how algorithmic routing reduced urgent triage requests from 25 to 10 per day while accelerating turnaround times.
Similarly, a report on workforce economics in an ACR Bulletin described how a large radiology practice balanced subspecialty backlogs against spare capacity. By implementing centralized scheduling and subspecialty pools, the group unlocked roughly 10% additional operational capacity without hiring extra personnel.
Remote access works only when accountability travels with it
Once worklists enable dynamic assignment beyond a single site, expanding allocation naturally leads to remote networks and teleradiology. However, expanding geographic reach alone does not solve clinical governance.
A study on a pediatric neuroradiology network demonstrated operational reach across 107 hospitals in 17 US states. Yet, clinical responsibility sets the true operational boundary. As noted in an RSNA review of remote radiology perspectives, licensing, credentialing, patient context, and professional-versus-technical billing splits remain strict constraints.
Furthermore, RCR census data warns that 90% of department heads worry that outsourced reports require local overreading, ultimately returning work to the originating team. Offloading a study remotely only works if patient context, legal accountability, and clear reporting chains travel alongside the image data.
The useful unit of allocation can be smaller than the whole study
To reconcile remote reach with strict accountability, diagnostic systems should shift their basic unit of allocation. Rather than forcing a full transfer of legal liability for an entire study, diagnostic infrastructure should support granular units of work.
Under this sub-task model, a complex measurement, subtle finding, or specialized second look is structured as a discrete subspecialty consultation. The primary signing radiologist retains full clinical context, overall case management, and final report liability, while remote experts provide targeted, bounded inputs.
This requires evolving from legacy systems that route:
Whole Study → Radiologist
to architectures routing:
Diagnostic Task → Expertise
Decoupling sub-tasks matches complex questions to subspecialists without disrupting local governance.
Radiology already relies on informal curbside consults and subspecialty referrals. The path forward requires converting these informal touchpoints into structured, traceable, and compensated workflows.
Leaders should start by mapping subspecialty backlogs, identifying where informal consults are occurring, and piloting bounded consultation workflows that track turnaround times and rates of interruption.
The radiologist shortage will not be solved by allocation alone. But before healthcare concludes that every capacity problem requires more people, it should ask a simpler question: how much of the expertise it already has is actually reaching the patients who need it?
The next generation of diagnostic infrastructure should not simply move studies between worklists. It should move expertise to where it creates the most clinical value.
