Deaths Linked to X-Rays and Radiotherapy Fell Sharply Over Two Decades, Yet Gaps Persist

Every year, tens of millions of Americans undergo imaging scans, interventional radiological procedures, or courses of radiation therapy, and for the vast majority these technologies are safe, routine, and lifesaving. Yet the machinery of modern radiology and radiation oncology is powerful enough that, in rare circumstances, complications arising from these procedures can contribute to a patient’s death. A new nationwide analysis has now put hard numbers on that risk across two decades of American medicine, and the picture it paints is broadly reassuring, with one important caveat: the benefits of safer practice have not been shared equally across the population.

The study, published in BMC Medical Imaging by Saad Arsalan Wasti of King Edward Medical University, Areesha Wasti of Continental Medical College, and Dori Woldu of Butaro Level Two Teaching Hospital, examined mortality associated with radiological procedures and radiotherapy in the United States between 1999 and 2019. Rather than relying on single-institution case series or voluntary reporting databases, the researchers turned to one of the most comprehensive mortality resources available: the Centers for Disease Control and Prevention’s WONDER Multiple Cause of Death database, which compiles every recorded death in the country along with the coded conditions and circumstances that contributed to it.

The methodological backbone of the study is the International Classification of Diseases, Tenth Revision, or ICD-10, the standardized coding system used on death certificates nationwide. Within that system, the code Y84.2 specifically designates deaths in which a radiological procedure or radiotherapy was recorded as either the underlying cause or a contributing factor. By searching for this code across the full 21-year window, the team identified a total of 22,542 deaths associated with radiological procedures and radiotherapy. That figure, spread across two decades and a population of more than 300 million, underscores how uncommon such deaths are, but the researchers were less interested in the raw count than in how the rate changed over time and who was most affected.

To make fair comparisons across years, the authors calculated both crude mortality rates and age-adjusted mortality rates per 100,000 population. Age adjustment is a critical statistical technique in epidemiology: because the American population has grown steadily older over the study period, a naive comparison of raw death counts could falsely suggest rising risk simply because there are more elderly people to begin with. The overall age-adjusted mortality rate across the study period was 0.33 per 100,000 population, with a confidence interval of 0.33 to 0.33, reflecting the enormous sample size and the precision that national-level data afford. In practical terms, for every 300,000 Americans, roughly one death certificate in a given year carried this radiological or radiotherapy-related code.

The headline finding is a clear and statistically robust decline. In 1999, the age-adjusted mortality rate stood at 0.50 per 100,000 population; by 2019 it had fallen to 0.24, a reduction of more than half. The researchers quantified this trajectory using Joinpoint regression, a technique that fits a series of joined straight lines to mortality data on a logarithmic scale and identifies the years at which the trend changes direction. From the fitted model they calculated the average annual percent change, or AAPC, over the whole period: minus 3.62 percent per year, with a 95 percent confidence interval running from minus 4.17 to minus 3.20 percent, and a P value below 0.000001. In plain language, the decline was steady, continuous, and almost certainly not a statistical fluke.

What drove this improvement? The study’s observational design cannot assign causation, but the timeline coincides with well-documented advances in radiation delivery and safety culture. Modern linear accelerators with intensity-modulated radiation therapy, image-guided targeting, and sophisticated treatment planning have progressively reduced the dose delivered to healthy tissue. In diagnostic and interventional radiology, dose-reduction protocols, better equipment regulation, and the growth of radiation safety programs have similarly tightened the margins of error. The steady, year-on-year nature of the decline, captured by the Joinpoint analysis, is consistent with a gradual accumulation of these safety improvements rather than any single turning point.

Yet the disaggregated data reveal a more complicated story. Men experienced higher mortality than women, at 0.38 versus 0.28 per 100,000 population, although both sexes showed significant declines over time with P values below 0.000001. Age was the most striking stratifier: mortality rose progressively with advancing age, peaking among adults aged 85 and older, whose crude mortality rate reached 2.94 per 100,000 population, nearly nine times the national average. This gradient likely reflects a combination of factors, including greater frailty, higher rates of cancer treatment in the elderly, and reduced physiological reserve for weathering procedural complications. It also serves as a reminder that procedure-related risk must always be weighed against the therapeutic benefit that imaging and radiotherapy provide to the very same older patients.

Geography and community type mattered as well. Rural areas recorded higher mortality rates than urban areas, at 0.38 versus 0.32 per 100,000 population, and the Midwest demonstrated the highest regional mortality rates among the census regions. The rural-urban gap is particularly notable because it echoes broader patterns in American health services research: rural hospitals often have older equipment, thinner staffing, and longer transfer distances when complications arise. The authors did not identify the mechanisms behind these disparities, but the finding aligns with a growing body of evidence that where a patient lives can shape the safety of the care they receive, independent of the procedure itself.

The study has limitations that the authors and readers should keep in view. Death certificate coding depends on the accuracy and consistency of the physicians and medical examiners who complete the certificates, and Y84.2 may be underused or inconsistently applied, particularly when a procedure-related death is obscured by the patient’s underlying illness. The database also cannot distinguish between diagnostic imaging, interventional procedures, and therapeutic radiotherapy, nor can it capture near-misses and nonfatal complications that never reach a death certificate. Because the analysis is ecological, it identifies associations at the population level rather than causal pathways in individual patients. These caveats do not undermine the central trends, but they do mean the absolute numbers should be interpreted as a floor rather than a ceiling.

Even so, the significance of the work lies in its scale and its framing. Most evidence on radiation-related harm comes from voluntary incident reporting systems, which are known to undercount, or from specialized registries covering narrow procedure types. A twenty-one-year, all-ages, all-states analysis provides a baseline against which future safety interventions can be measured. The authors suggest their findings may help inform future studies evaluating procedural safety and healthcare disparities, and the disparity map they have drawn, older adults, men, rural residents, and residents of certain regions, gives researchers and policymakers concrete places to look. For a field that touches nearly every cancer patient and countless diagnostic workups, knowing that mortality has fallen by half while identifying who remains at elevated risk is exactly the kind of evidence that turns good intentions into targeted policy. The machines that peer inside our bodies and burn away our tumors have never been safer; the task now, this study implies, is to make sure that safety reaches every corner of the country.

Subject of Research: National trends and demographic and geographic disparities in mortality associated with radiological procedures and radiotherapy in the United States from 1999 to 2019

Article Title: National trends and disparities in mortality associated with radiological procedures and radiotherapy (1999–2019)

Article References: Wasti, S. A., Wasti, A., & Woldu, D. (2026). National trends and disparities in mortality associated with radiological procedures and radiotherapy (1999–2019). BMC Medical Imaging. https://doi.org/10.1186/s12880-026-02783-5

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02783-5

Keywords: radiotherapy, radiological procedures, mortality trends, health disparities, CDC WONDER, ICD-10, Joinpoint regression, age-adjusted mortality, rural health, radiation safety, epidemiology, BMC Medical Imaging