Nine Takeaways from New Article Examining Health Equity in the Radiology Field

In a provocative new article, radiology researchers discuss the impact of social determinants of health (SDoH) upon access to care and patient outcomes, and present strategies within the realms of radiology education, research, clinical care, and innovation that may help mitigate health-care disparities.

In a closer look at the dilemma of health care disparities with timely and appropriate access to imaging, radiology researchers have authored a new publication that examines ongoing challenges in the field, the broader impact of social determinants of health (SDoH), and potential practical solutions to advance health equity.

Here are nine key takeaways from their article, which was recently published in the American Journal of Roentgenology.

1. Trust of healthcare institutions is a significant issue among multiple racial and ethnic groups. The article authors explained that an emphasis on building long-term relationships in these populations may help facilitate more diverse representation in study cohorts with perhaps improved generalizability of study findings as a result.

2. In multiple studies, researchers noted increased lung cancer screening as well as appropriate follow-up after abnormal mammograms with the use of patient navigator programs at community health centers and in populations that have limited access to medical care.

3. In order to facilitate increased access to imaging modalities, researchers suggested the formation of transdisciplinary collaborations between radiologists, patients, community members and clinicians from other disciplines could go a long way toward bolstering lung cancer screening in high-risk populations. The article authors noted these collaborations led to a successful low-dose CT lung cancer screening initiative in rural areas in 2021 and a 2021 collaboration between radiology, primary care and psychiatry helped overcome fragmented care and facilitated lung cancer screening in patients with mental illness.

4. When making recommendations between comparable imaging studies in line with appropriateness criteria from the American College of Radiology (ACR), the article authors emphasized increased cognizance and consideration of out-of-pocket costs incurred by patients as well as potential socioeconomic constraints that may factor into imaging decisions.

(Editor’s note: For related articles, see “Medicare Mammography Study Shows Black Women Had Less Initial Access to Imaging Advances than White Women,” “What a New MRI Study Reveals About Brain Aging and Racial Disparities” and “Study Finds Disparities with Follow-Up After Incomplete Mammography Exams.”)

5. Social determinants of health (SDoH), ranging from socioeconomic factors and cultural identity to race and geographic proximity to health care institutions, reportedly have a cumulative effect of up to 50 percent on variation with clinical outcomes.

6. Citing a curriculum review of medical schools by the American Association of American Medical Colleges (AAMC), researchers noted that 87 medical schools addressed SDoH in the first year of medical school but less than half of those colleges tackled the topic in the second year of medical school.

“As medical students transition to radiology residency, this gap in medical school curricula will potentially perpetuate existing health disparities in radiology,” wrote article co-author Efren J. Flores, M.D, an associate professor of radiology at Harvard Medical School, and colleagues.

7. Emphasizing an ongoing need for a more diverse makeup of the radiology workforce, the researchers noted that among radiology trainees, less than 1 percent identify as Pacific Islanders, Native Americans or Native Hawaiians, 4 percent are Black or African American, and 6 percent identify as Hispanic or Latino. While women comprise more than 50 percent of graduating medical trainees, the article authors pointed out that women currently comprise 29 percent of the radiology workforce.

8. Offering cautionary notes about artificial intelligence (AI), the researchers note the use of non-representative data in the training of AI models may perpetuate bias. They maintained that diverse, high-quality databases are required for explainable AI models and may help bolster transparency in regard to avoiding potential disparities. The article authors also maintained that ongoing evaluation and adjustment of AI models is necessary in order to mitigate potential bias.

9. Incorporating the perspectives of diverse communities may play a significant role in facilitating “equitable implementation” of novel technologies in the realm of radiology, according to the researchers. They suggested that theragnostic molecular imaging, AI enhancement of language translation services and the ongoing advent of radiomics in oncology care are potentially transformative technologies that would benefit from diverse stakeholder input.

“If leveraged adequately, these novel tools can serve as vehicles to advance health equity,” noted Flores and colleagues.