Radiology’s Perfect Storm Around Workforce Gaps + an Aging Population

Radiology is entering a perfect storm. By 2030 roughly one in five Americans (~70 million people) will be over age 65, and seniors already consume about 30 percent of all imaging studies. At the same time, hospitals nationwide are struggling to fill radiology jobs. Radiologists have left the workforce at a 50-percent-higher rate since the start of the pandemic. Right now, there are about 2,000 radiologist positions posted on the American College of Radiology job board, reflecting a shortage that has left many departments short-staffed. The shortage, combined with a rising demand, can force imaging delays measured from days to upwards of two weeks.

Finally, artificial intelligence is flooding practices with new findings. Hundreds of FDA-cleared AI tools now analyze X-rays, CTs and MRIs, with over 500 approved radiology algorithms. These algorithms can catch subtle disease—but they also flag a torrent of incidental findings, exacerbating the follow-up backlog and overwhelming care coordinators.

Heightening demand and shrinking workforce

Radiology demand is surging faster than the workforce. U.S. Census and Neiman Institute projections show the oldest cohorts exploding: by 2055 Americans aged 85–94 will grow nearly 150 percent, and those 95+ will nearly quadruple. The Neiman studies estimate imaging utilization rising roughly 17–27 percent by 2055 (depending on modality), due in large part to this aging boom. In other words, current training pipelines would barely keep pace with demand, and the existing radiologist shortage is projected to persist without intervention. Many imaging centers have managed the gap only by creative workarounds (teleradiology, driving wages up, etc.), but these are stopgaps at best. The endgame? Sicker patients are waiting longer for diagnosis.

AI’s Explosion of Incidental Findings

AI tools are now rapidly deployed across radiology workflows. Hundreds of algorithms effectively detect lung nodules, strokes, fractures, vessel aneurysms, and more, often in parallel on each scan. By raising the analytic sensitivity of every exam, AI inevitably increases the number of incidental discoveries. A CT scan intended for one complaint may also pick up unexpected nodules, mass lesions, or early cancers.

On the one hand, it’s clearly a positive to be able to catch disease earlier. But it also means radiologists and care teams are now generating far more follow-up work than before. Every additional nodule or lesion noted in a report translates into a cascade of care. In a fully-resourced system, this could improve outcomes. But in practice, the efficiencies created by AI collide with the workforce gap: there are fewer hands to act on each alert. In effect, AI is amplifying downstream demand for care coordination and putting more strain on an already overloaded system.

The follow-up gap: patient safety and liability risks

Without systematic processes, this creates a dangerous bottleneck. Even before AI, studies have shown that a large fraction of radiology follow-up recommendations simply fall through the cracks. In one analysis, up to 10 percent of imaging reports carried a follow-up recommendation, yet about half of those advised exams were never done. When patients fail to get that follow-up scan, malignancies can progress undetected, creating worse outcomes for patients.

This gap in care also raises liability. An undetected or uncommunicated actionable finding can lead to malpractice claims if a patient is harmed. With more incidental lesions being found by AI, the stakes are even higher.

Building high-reliability follow-up systems

The solution to addressing this perfect storm lies in system-level innovation. We must transform actionable-finding management into a high-reliability process with automation and clear handoffs. An ideal system begins with AI and natural language to identify recommendations for additional imaging and drives automated care orchestration through to completion. The solution is not a stand-alone technology. It is the blend of process enablement, people empowerment, and technology; technology that empowers the way clinicians work and reduces risk across the patient pathway by closing gaps and mitigating failure points.

Key components of a high-reliability follow-up system include:

Ø  Automated identification of follow-ups. NLP or AI tools scan every report for any recommendation or actionable finding (nodule, mass, aneurysm, etc.) and create smart worklists based on the patient, the follow-up, and the provider’s judgment on the next best step.

Ø  Automated care orchestration. To best manage the high volume of follow-ups, automation must alleviate manual and repetitive tasks, provide for automated and highly configurable communication pathways, and only escalate to the care team when needed. Technology should be the force multiplier that elevates humans to work at the top of their capabilities by eliminating activities that don’t require their emotional or clinical skills.

Ø  EHR-integrated workflows. Flagged findings auto-generate tasks or alerts for the referring physician and provide for completion tracking all within the established EHR workflow. The right solution only sends an alert when needed to avoid noise and distraction.

Ø  Patient engagement automation. Instead of relying on a doctor’s memory or a patient’s initiative, the system sends automated reminders (calls, letters, portal messages) to schedule the imaging.

Ø  Continuous monitoring and feedback. Advanced analytics track follow-up completion rates, escalation rates, and the root causes of failure points to support ongoing improvement. The right solution empowers evergreen quality improvement that is actionable and meaningful.

Crucially, these measures do not mean more work for care teams, radiologists or PCPs. In fact, by automating the tedious tracking and outreach, they allow clinicians to focus on care, not on paperwork. Healthcare leaders must act now. The coming decade will see more scans done on older, sicker patients and more incidental findings identified than ever before. If we build the right infrastructure, built on high-reliability principles, we can meet this challenge. Otherwise, every AI advance risks amplifying our systemic weakness. By investing in integrated follow-up management today, health systems can turn AI into a patient-safety boon rather than a liability.

Angela Adams, R.N., B.S.N., AALC, has been advancing the industry by applying AI to improve healthcare outcomes for over a decade. Angela started her career as a critical care medicine nurse at Duke University Medical Center. During her time in the hospital setting, Angela became increasingly frustrated with the inefficiencies in patient care. Driven to make a broader impact, Angela looked to the emerging healthcare AI segment for solutions that would allow her to help patients as well as assist clinicians to become more effective and efficient in solving complex medical issues. She helped advance AI adoption and overcome skepticism at companies like Jvion (acquired by Lightbeam Health Solutions), where she applied deep machine learning to lower nosocomial event rates and prevent patient deterioration. She went on to create her most recent solution at Inflo Health, where she focuses on missed follow-up radiology appointments.