In healthcare, more isn’t always better. Overuse, tests and treatments that offer little or no clinical benefit, remains a persistent and costly challenge. In fact, estimates suggest that up to 30% of U.S. healthcare spending goes toward unnecessary care.
As artificial intelligence (AI) and automation make it easier to detect findings and flag next steps, the potential for unintended overuse also rises. What’s needed is not just more detection—but more discretion. Ensuring the right follow-up happens at the right time, for the right patients, requires more than algorithms. It requires a thoughtful, human-centered approach that prevents overuse without missing what matters.
The Problem with Overuse
Overuse, defined as care where potential harms outweigh the expected benefits, is widespread, especially in diagnostic imaging and specialty referrals. And it’s not just a theoretical concern. Overuse contributes to patient anxiety, unnecessary procedures, incidentalomas that trigger cascades of testing, and ballooning healthcare costs.
Choosing Wisely, a national initiative led by the ABIM Foundation, has been a leading voice in this conversation. Since 2012, more than 80 specialty societies have released over 500 recommendations to reduce unnecessary tests and treatments. Despite growing awareness, physicians still report feeling pressure to “cover their bases,” avoid malpractice risk, and meet patient expectations, even when those expectations run counter to clinical guidelines.
Radiology follow-up is a uniquely sensitive area. While follow-through on incidental findings can be lifesaving, reflexively acting on every finding without context or confirmation may lead to low-value care. A measured, clinician-led approach is essential.
Three Principles for Preventing Overuse in Follow-Up
High-reliability approaches to radiology follow-up can strike the right balance, ensuring timely care for those who need it while avoiding unnecessary intervention for those who don’t. The following three principles are essential to preventing overuse:
1. Centering Clinical Judgment
Automated detection is only useful if it supports clinical decision-making. That’s why high-quality follow-up systems place clinicians at the center of the process. When a recommendation for follow-up is identified in a radiology report, it should be:
- Simplified for clarity, especially when phrased ambiguously in free text.
- Verified when clinical uncertainty exists—leveraging secondary review or clinical triage to avoid unnecessary cascades.
- Directed back to the referring provider with an opportunity to confirm or defer follow-up based on the full clinical picture.
Physician leadership is central to this process. Studies from the Choosing Wisely campaign show that when clinicians are engaged and supported, they are more likely to reduce unnecessary tests. Radiology follow-up efforts that align with this ethos promote safety without undermining autonomy.
2. Empowering Patients with Clarity and Choice
Patient empowerment is a safeguard against overuse, but only if patients are informed in a way that builds trust. Effective follow-up processes communicate:
- Why a follow-up is (or isn’t) needed.
- What the next step entails.
- How to access support, ask questions, or opt out if appropriate.
This kind of patient engagement is especially important for historically marginalized communities, where trust in the healthcare system may be lower and financial harm from unnecessary care more acute.
Focus groups conducted by the ABIM Foundation show that patients want to be heard, and are more likely to accept alternatives to requested tests if clinicians explain the rationale and offer a clear care plan.
Avoiding overuse doesn’t mean saying “no.” It means saying, “Let’s talk about what’s best for you.”
3. Creating Accountability and Transparency
Preventing overuse requires more than good intentions. Health systems need transparent processes that document:
- What follow-up was recommended.
- Whether it was completed, declined, or deemed unnecessary.
- Who made the decision and how it aligns with clinical standards.
This level of accountability supports continuous quality improvement, protects against medical-legal risk, and builds a defensible record of thoughtful, patient-centered care.
Organizations that embed these safeguards report fewer missed findings and less unnecessary imaging, all while improving clinician confidence in their follow-up workflows.
Aligning with Choosing Wisely: In Practice, Not Just Principle
Choosing Wisely emphasizes conversation over command. Rather than blanket rules, it encourages clinicians and patients to collaborate around appropriateness. This collaboration can be supported at scale through agentic AI solutions and a high-relability approach to follow-up care.
Radiology follow-up is an ideal test case. By combining AI-enabled high-reliability, evidence-based triage, clinician validation, and clear patient communication, health systems can:
- Reduce false positives and follow-up fatigue.
- Avoid duplication of tests and unwarranted referrals.
- Increase adherence to necessary care while decreasing unnecessary utilization.
It’s not just possible to prevent overuse in radiology follow-up, it’s necessary. The downstream risks of inaction are too high, and the opportunity to do better is within reach.
The Future of Radiology Follow-Up: Smarter, Safer, More Human
Radiology is evolving. AI is accelerating detection. Guidelines are becoming more nuanced. But the fundamentals remain the same: delivering the right care to the right patient at the right time.
That includes not delivering care that isn’t needed.
As health systems confront rising scan volumes, incidental findings, and staffing constraints, the path forward lies in systems that reinforce human expertise, not override it. Follow-up protocols that reflect the lessons of Choosing Wisely—clinical leadership, patient partnership, and defensible processes—can reduce harm, build trust, and deliver higher-value care.
In the end, the solution to overuse isn’t more technology or more tests. It’s more wisdom in how we use them.
