In 2009, a scandal erupted at Cedars-Sinai Medical Center in Los Angeles when a computed tomography (CT) programming error caused more than 200 patients to receive severe radiation overdoses.
The problem had gone undetected for a year and a half until a patient complained about losing hair after a scan.
Although egregious, it wasn’t an isolated event.
Patients often get two or three times more radiation than they need during a CT scan; some receive much more excessive doses, according to University of California San Francisco (UCSF) radiologist and epidemiologist Rebecca Smith-Bindman, M.D.
Yet the public is still largely unaware of this lack of standardization in radiation use. That’s distinct from other CT safety problems of unnecessary testing and false positives.
The Centers for Medicare & Medicaid Services (CMS) has unveiled a quality measure to prod providers to reduce excessive radiation use. It can also help journalists to spotlight the problem.
The measure, which Smith-Bindman’s team developed, will be phased into CMS quality programs over several years. The first voluntarily reported data, for procedures performed in 2025, will become public in 2026. (See details, below.)
Meanwhile, journalists can educate the public about the importance of monitoring their imaging.
“Patients should let their physicians and all healthcare providers know that optimizing radiation dose is important to them,” Smith-Bindman said via email, adding, “The more people who know about this issue and insist on improving the safety of CT, the more we can standardize the use of radiation in CT imaging.”
Saving lives
For decades, safety experts have expressed unease about the escalating use of CT, a speedy and reliable diagnostic tool that often eliminates the need for hospitalization and exploratory surgery. Various estimates indicate that 80 to 90 million scans are performed annually in the U.S.
The technology has a downside of emitting ionizing radiation that can damage human DNA. Studies have linked that radiation to a small risk of certain cancers, which amounts to thousands of cases a year.
Implementing the radiation dose measure is expected to prevent about 14,000 cancers among Medicare beneficiaries annually and save the Medicare program $1.86 billion to $5.21 billion in annual cancer treatment costs.
Benefits will also accrue to non-Medicare patients, particularly children, who usually have more time to develop cancer following radiation exposure.
The measure sets lower and upper thresholds based on patient characteristics and type of CT scan. It’s expressed as the percentage of time that an excessive dose is used or an image has poor quality.
While modifying doses may seem simple, Smith-Bindman noted that hospitals each have hundreds to thousands sets of instructions, or protocols, that technologists use to program scanners for different clinical indications. Each protocol needs to be reviewed and potentially revised.
Some good news is that a randomized trial at 100 hospitals and outpatient radiology practices showed that providing detailed feedback and educating radiologists and technologists as part of a quality improvement effort can reduce high-dose scans.
Transparency takes time
The new measure will allow journalists to report on how well individual hospitals and outpatient providers manage radiation dosing and compare their performance to other facilities.
Obtaining facility-specific data on CT radiation use can be challenging, although some journalists have managed to do it by interviewing patients who’ve been harmed.
The Leapfrog Group, a quality watchdog group, collects data on pediatric CT radiation doses for head and abdomen/pelvis scans in its voluntary hospital survey.
Easier to come by have been data on how often facilities order unnecessary scans. For example, in 2011 CMS implemented a set of quality measures that were designed to reduce duplicative CT scans. (For a great example of reporting on one hospital’s duplicative use of scans, see sidebar.)
The new measure underwent rigorous vetting and garnered broad support from patients and safety organizations as well as some radiologists.
Its electronic format is meant to ease reporting for providers, and a company formed by UCSF researchers, Alara Imaging, will issue free education materials and software for formatting radiology data.
Despite those pluses, not everyone is happy.
The American College of Radiology, for example, complained about the amount of work involved for hospital staff to implement the measure. It also disputed its methodology and the term “excessive radiation dose” in the title.
Such provider pushback is one reason that quality measures often take years to implement, and the CT radiation measure is no exception.
Hospitals won’t be required to report outpatient data until 2027, following voluntary reporting in 2025 and 2026 via the Hospital Outpatient Quality Reporting Program.
In 2025, an inpatient version will be included on a list of 12 optional electronic clinical quality measures (eCQMs), of which hospitals must pick three to report. That year, hospitals also will be required to report three eCQMs: safe use of opioids, caesarian birth rate, and severe obstetric complications.
That data will be released in 2026 via CMS’s Provider Data Catalog site.
Regarding the inpatient CT radiation measure, CMS said via email that it intends “to make additional eCQMs mandatory in the future” but provided no timeframe.
Essential reading
- Quality measure: Excessive Radiation Dose or Inadequate Image Quality for Diagnostic Computed Tomography (CT) in Adults
- Study on variability in CT dosing across hospitals
- Randomized trial on interventions to lower radiation use
- Federal Register description of the development of the excess radiation measure, including CMS responses to comments
Sidebar:
Exposing double scans
Notable reporting on a hospital’s use of CT scans was a 2011 story by Lisa Chedekel, co-founder of the nonprofit Connecticut Health Investigations Team (C-HIT).
Her scrutiny of CMS quality data showed how UConn John Dempsey Hospital in Farmington used “combination” scans at ten times the national rate. Instead of a single scan, patients got two — one with contrast material and one without.
Chedekel died in 2018, and her partner in C-HIT, Lynne DeLucia, retired and shuttered the project last year.
DeLucia said in an email that readers had “an insatiable appetite” for quality stories such as the one about combination scans and wanted to know how local health care providers stacked up statewide and nationally.
“In telling these stories, we tried to combine data and real-life experiences, to bring a deep understanding of the issues and to empower people to make smart health care choices,” she said.
