FDA recalls for radiation oncology devices have remained stable from 2015 to 2025 despite the rapid adoption of AI and software-based technologies, with external beam systems accounting for over 60% of all radiation oncology-related recalls, mostly classified as class II severity.
- Of 30,648 total FDA recalls analyzed from 2015-2025, only 426 (1.39%) were related to radiation oncology, showing this device category represents a small fraction of overall medical device recalls.
- External beam systems accounted for 61.6% of radiation oncology recalls, while planning and software systems made up 34.4% of these incidents.
- Software issues were the root cause in 45% of recalls, exceeding 50% in six of the eleven years studied, making it the most common failure mechanism.
- 97.7% of radiation oncology recalls were classified as class II, indicating temporary or reversible health consequences rather than serious or life-threatening outcomes.
- Despite 1,357 AI-enabled devices entering the market since 2018, annual recall counts showed no significant temporal trend, demonstrating stability in recall frequency despite increasing technology complexity.
Recall frequency for radiation oncology devices approved by the U.S. Food and Drug Administration (FDA) remains stable despite the rapid adoption of software-based and AI-powered technologies.
External beam systems accounted for over 60% of all radiation oncology-related FDA recalls, with most recalls being class II, wrote Mariam Rowe, MD, from Royal Brisbane and Women’s Hospital in Queensland, Australia, and Dylan Rowe from Griffith University in Brisbane, Queensland, Australia.
“These findings underscore the need for continued post-market surveillance, rigorous commissioning and revalidation, lifecycle oversight, and clinician education,” Rowe and Rowe wrote in Practical Radiation Oncology.
Post-market surveillance of radiation oncology devices is an important part of ensuring device quality and patient safety. Prior research suggests that these devices have been recalled at higher rates than many other medical device categories. Software-related problems accounted for many recalls in this research.
And since 2015, the radiation oncology field has adopted more AI and machine learning-powered tools for auto-segmentation, treatment planning, and image guidance.
Rowe and Rowe analyzed radiation oncology device recalls from 2015 to 2025 (n = 30,648 recalls) to find out whether this has resulted in recall increases. They focused on device category, recall severity, FDA-determined root cause, and quantity in commerce.
Of the total recalls during the study period, 426 (1.39%) were related to radiation oncology. External beam systems accounted for 61.6% of the radiation oncology-related recalls, and planning/software systems made up 34.4% of these recalls.
“Software issues were the most common root cause [45%], exceeding 50% in six of 11 years,” the researchers wrote.
Most recalls were class II (97.7%). These indicate potential temporary or reversible health consequences.
From 2018 onward, 1,357 AI-enabled devices entered the market, with 70 (5.16%) identified as radiation oncology-related. Over this period, 68 AI-enabled radiation oncology devices were recalled, with 62 (91.2%) involving a single planning system.
The study authors highlighted their results as providing “a contemporary benchmark” to inform patient safety strategies. They also found that annual radiation oncology device recall counts showed no significant temporal trend during the study period. This points to stability in recall frequency despite the technology becoming more complex.
“Together, these findings suggest that although recall incidence has plateaued, the underlying mechanisms driving device failure have not been eliminated,” the authors wrote.
Read the full study here.
