Radiology and Imaging Misreads: When the Scan Showed It and the Report Missed It

The Finding Was on the Film. The Report Said Normal.

A radiology misread is a diagnosis that existed and went unspoken: the nodule on the chest film, the fracture on the X-ray, the mass on the MRI, present in the pixels and absent from the report.

Peer-reviewed medicine puts the day-to-day radiologist error rate at roughly 3 to 5 percent of studies.

Multiply that by millions of scans and the missed finding becomes one of the most common ways serious disease escapes early treatment.

Not every miss is malpractice.

The law asks whether a reasonably careful radiologist would have caught it and reported it, and the images themselves preserve the answer.

Call (888) 713-6653 for a free review; the scan that was misread the first time can be read again.

How Often Scans Get Misread: What the Medical Literature Actually Shows

Radiology's own journals have measured the problem for decades. The day-to-day, real-time error rate averages 3 to 5 percent of studies, and when researchers re-examine imaging with the benefit of hindsight, retrospective discrepancy rates average around 30 percent.^[1] Those two numbers describe different things, a working radiologist's misses versus what a second look can find, and both matter in litigation.

The lung cancer data is the starkest. A clinical-practice study found 19 percent of lung cancers were missed on chest radiographs, at a median size of 16 millimeters, with a median diagnostic delay of 472 days.^[2] In the Mayo screening program, 90 percent of peripheral lung cancers found on screening films were visible in retrospect on earlier ones. A tumor that grows for fifteen extra months is a different disease by the time someone finally names it.

The Four Ways a Radiology Read Goes Wrong

A landmark review of 656 delayed radiology diagnoses sorted the failures, and four patterns carry most cases:^[3]

• Under-reading, 42 percent of errors. The finding is present and simply missed. The most common failure, and the one a re-read exposes immediately.
• Satisfaction of search, 22 percent. The radiologist finds one abnormality and stops looking, missing the second. The rib fracture gets reported; the lung nodule behind it does not.
• Misjudgment. The finding is seen and called benign, or dismissed as artifact, when its features warranted workup.
• Communication failure. The finding is seen, reported, and never reaches anyone who acts on it, a category with its own section below because it is its own kind of case.

When a Miss Becomes Malpractice, and Why Reading Speed Enters the Case

The legal test is the same one governing every provider: what a reasonably careful radiologist would have done with that study. Some subtle findings are missable by careful readers, which is why honest case screening starts with an independent re-read by a qualified radiologist rather than a lawyer's opinion of a shadow.

Workload evidence increasingly matters. Researchers calculated that meeting modern volume demands requires interpreting an image every three to four seconds across a full shift, a pilot study found major-error rates rose from 10 percent to 26.6 percent when radiologists read at double speed, and an analysis of 2.9 million teleradiology reads found errors clustering in the tenth through twelfth hours of shifts and on the highest-volume shifts.^[4] When a group builds a production line that outruns careful reading, the negligence can be institutional as well as individual.

The Correct Report Nobody Acted On Is Still a Case

The American College of Radiology's communication standards require more than filing a report: critical and unexpected findings call for direct, documented communication to the treating clinician in time to matter. A finding that dies in an unread report can breach the standard even though the read itself was perfect.

Follow-up recommendations fail at rates that surprise everyone but the researchers. In one study of incidental lung nodules, only 29 percent of patients received the follow-up imaging the radiologist had recommended in the report.^[5] Responsibility for those failures spreads across the ordering physician, the practice's tracking systems, and sometimes the radiologist's manner of flagging, and the electronic record usually shows exactly where the handoff broke. Patient portals have added a timestamp to all of it, and our page on getting your complete records and imaging explains how to capture that trail.

Who Answers for a Misread: the Radiologist, the Group, and Sometimes the Hospital

Radiologists are among the most-sued physicians in America, and the reason is the stakes of what they miss: in a study of more than 8,000 radiologists, 31 percent had faced at least one claim, with failure to diagnose the leading allegation and missed breast cancer the most common missed diagnosis.^[6]

The defendant list takes work, because hospital radiology is heavily outsourced. The read may belong to a private group, a national teleradiology company, or a contractor the hospital holds out as its own, and the hospital's exposure runs through the agency rules covered on our page about hospital liability for contractor physicians. Causation then has to be built with oncology and specialty testimony showing what earlier detection would have changed, the same architecture as any cancer misdiagnosis claim, and it all runs inside deadlines built around discovery rules that start clocks earlier than patients expect.

One currency note: artificial intelligence now assists a large share of reads, with more than a thousand FDA-authorized radiology AI tools on the market. Nearly all are assistive. The interpreting radiologist remains the physician of record, and responsibility for the final read has not moved.