Computed tomography (CT) scans may account for 5% of all cancers annually, according to a new study out of UC San Francisco that cautions against overusing and overdosing CTs.
The danger is greatest for infants, followed by children and adolescents. But adults also are at risk, since they are the most likely to get scans. Nearly 103,000 cancers are predicted to result from the 93 million CT, also known as CAT, scans that were performed in 2023 alone. This is 3 to 4 times more than previous assessments, the authors said.
Reducing the number of scans and reducing doses per scan would save lives.”
REBECCA SMITH-BINDMAN, MD
The study, which was funded by the National Institutes of Health, appears April 14 in JAMA Internal Medicine.
“CT can save lives, but its potential harms are often overlooked,” said first author Rebecca Smith-Bindman, MD, a UCSF radiologist and professor of epidemiology and biostatistics and obstetrics, gynecology and reproductive sciences.
“Given the large volume of CT use in the United States, many cancers could occur in the future if current practices don’t change,” said Smith-Bindman, who is also a member of the Philip R. Lee Institute for Health Policy Studies and directs the Radiology Outcomes Research Lab.
“Our estimates put CT on par with other significant risk factors, such as alcohol consumption and excess body weight,” she said. “Reducing the number of scans and reducing doses per scan would save lives.”
Benefits and potential dangers
CT is both indispensable and widely used to detect tumors and diagnose many illnesses. Since 2007, the number of annual CT exams has surged by 30% in the U.S.
But CTs expose patients to ionizing radiation – a known carcinogen – and it’s long been known that the technology carries a higher risk of cancer.
To assess the public health impact of current CT use, the study estimates the total number of lifetime cancers associated with radiation exposure in relation to the number and type of CT scans performed in 2023.
“Our approach used more accurate and individualized CT dose and utilization data than prior studies, allowing us to produce more precise estimates of the number of radiation-induced cancers,” said co-author Diana Miglioretti, PhD, a breast cancer researcher and division chief of biostatistics at UC Davis. “These updated estimates suggest the excess risks – particularly among the youngest children – are higher than previously recognized.”
Researchers analyzed 93 million exams from 61.5 million patients in the U.S. The number of scans increased with age, peaking in adults between 60 to 69 years old. Children accounted for 4.2% of the scans. The researchers excluded testing in the last year of a patient’s life because it was unlikely to lead to cancer.
Future cancers from radiation exposure
Few patients and their families are counseled about the risk associated with CT examinations …”
MALINI MAHENDRA, MD
Adults 50 to 59 had the highest number of projected cancers: 10,400 cases for women, 9,300 for men. The most common adult cancers were lung, colon, leukemia, bladder and breast. The most frequently projected cancers in children were thyroid, lung and breast.
The largest number of cancers in adults would come from CTs of the abdomen and pelvis, while in children they came from CTs of the head. Projected cancer risks were highest among those who underwent CT when they were under 1 year old. They were 10 times more likely to get cancer compared to others in the study.
The researchers said some CT scans are unlikely to help patients, and are overused, such as those for upper respiratory infections or for headaches without concerning signs or symptoms. They said patients could lower their risk by getting fewer of these scans or by getting lower dose scans.
“There is currently unacceptable variation in the doses used for CT, with some patients receiving excessive doses,” Smith-Bindman said.
Co-author Malini Mahendra, MD, a UCSF assistant professor of Pediatric Critical Care, said it was important that families understand the risk of developing cancer from pediatric scans.
“Few patients and their families are counseled about the risk associated with CT examinations,” she said. “We hope our study’s findings will help clinicians better quantify and communicate these cancer risks, allowing for more informed conversations when weighing the benefits and risks of CT exams.”
This is academic fear-mongering, plain and simple. The estimates are modeled, not observed, and ignore real-world evidence - like the absence of elevated cancer rates in pilots and flight crews exposed to similar or higher radiation levels. Claiming CT scans cause 5% of all cancers is like calling air travel a major carcinogen. It’s spreadsheet sensationalism, not serious clinical risk analysis. Garbage in, garbage out.
It’s very interesting, and the corresponding author looks well qualified and credible at first glance. And it’s really important not to bullshit ourselves or try to ignore bad news. I think CT scans are great and I’ve never worried about radiation, so this would be bad news for sure if true.
I’m curious what they say about doses and being exposed to excessive doses during scans. I’m not sure how to shop around on that point. Like, how would I know what machine settings the doctor is using?
However, I am always cautious of these sorts of articles where they say things are over-used. It’s usually politically or financially motivated. When they mention people with headaches, I just can’t imagine a doctor throwing someone into a CT for a random headache. However, a patient with persistent headaches, even absent other symptoms, a CT seems reasonable IMO.
So, your chance of getting cancer from a scan is .0011%.?
0.11%… You’re off by a factor of 100.
But that is still only a little over 1 in a thousand. I guess everyone has to weigh their risks. I am not a fan of CT scans. I just pretend I got a bad result and take appropriate measures like keeping my lipid scores very low.
Well, I guess there’s a difference between screening vs diagnostic use. For example, if you have cancer, CT is one of the most powerful ways to detect it while small and operable. It would be silly to refuse a scan (IMO) if the doctor thinks it’s necessary to rule out something sinister.
Edit: for context, isn’t is roughly stated that 1/3 to 1/4 people will get cancer in their lifetime anyway? So the default state of “do nothing” often results in cancer, and your base risk isn’t 0 without CT. In that case, a small amount of extra risk to potentially detect it early is a different risk calculation.