Access to radiology specialists in cardiothoracic imaging, breast imaging and nuclear medicine may become more challenging for patients in light of new research on attrition rates in radiology.
For the retrospective study, recently published in the American Journal of Roentgenology, researchers reviewed data from a total of 159,490 radiologist years for 29,770 subspecialist radiologists over an eight-year period from 2014 to 2022.
The study authors found an overall attrition rate of 2.2 across radiologist subspecialties but noted higher attrition rates for cardiothoracic imaging (4.3 percent), nuclear medicine (2.9 percent) and breast radiologists (2.8 percent).
Employing abdominal radiology as the reference subspeciality for a multivariable logistic regression analysis, the researchers noted 81 percent and 31 percent higher likelihoods of attrition for cardiothoracic imagers and breast radiologists respectively.
In addition to potentially exacerbating challenges in access to these subspecialists during non-business hours and in certain geographic regions, the study authors maintained that these ongoing trends in subspecialist attrition could have a compounding effect on other radiologists.
“Moreover, relatively higher attrition in certain subspecialties (as presently observed for breast imaging and cardiothoracic imaging based on adjusted odds) could, over time, lead to disproportionate shortages in those subspecialties. This effect could lead to a vicious cycle where a relatively pronounced shortage in a specific subspecialty abruptly heightens workload demands and call burden on that subspecialty’s remaining radiologists, further exacerbating workforce departures,” posited lead study author Eric Christensen, Ph.D., the director of economic and health services research at the Harvey L. Nieman Health Policy Institute in Reston, Virg., and colleagues.
The researchers also noted that mean estimated career length was higher for academic radiologists in contrast to those in non-academic settings, a finding that held true for women (37.3 years vs. 33.6 years) and men (38.4 years vs. 35.1 years.)
Three Key Takeaways
- Higher attrition in certain subspecialties. While the overall radiology subspecialty attrition rate was 2.2 percent, it was significantly higher in cardiothoracic imaging (4.3 percent), nuclear medicine (2.9 percent), and breast imaging (2.8 percent).
- Risk of future workforce shortages. Higher attrition in these subspecialties could reduce patient access and increase workload for remaining specialists, potentially creating a cycle that further drives departures.
- Practice setting differences matter. Academic radiologists tend to have longer career lengths than non-academic radiologists, suggesting factors such as workload expectations, compensation structure, and burnout may influence attrition trends.
The study findings also revealed that breast imaging had the lowest mean estimated career length among subspecialties for men (35.7 years) and women (35.3 years) in academic settings. Cardiothoracic imaging was associated with the lowest mean estimated career length for men (31.1 years) and women (30.9 years) in non-academic settings, according to the study authors.
“Differences in attrition between non-academic and academic radiologists could relate to differences between the two practice environments in productivity expectations, compensation, salary stability, team environment, and burnout rates,” suggested Christensen and colleagues.
(Editor’s note: For related content, see “New Study Examines Turnover Trends in Radiology,” “Emerging Trends with Radiology Practice Closures Point to Increased Subspecialization” and “Radiology Workforce Study Finds Greater Attrition of Women, Subspecialists and Non-Academic Radiologists Over Eight-Year Period.”)
In regard to study limitations, the authors acknowledged that they did not examine causes of the varied attrition rates with radiology subspecialties. The researchers also conceded that onsite, remote and hybrid work arrangements were not factored into the study. They noted that pediatric radiology was not evaluated in the study due to data limitations.
