Online information about AI in mammography is written at a level that may be hard for many women to understand, suggest findings published March 10 in Clinical Breast Cancer.
The average readability across webpages found via AI-related internet search terms in mammography is at a college level, reported a team led by Tanvin Araen and colleagues from Brown University. This far exceeds national recommendations for readability in medical materials.
“Because patients are likely to encounter both online patient education materials and [non-online patient education materials] sources when searching online, clinicians should be prepared to guide them toward accessible, reliable resources,” the Araen team wrote.
As AI continues to be integrated into breast imaging, including for breast cancer screening, women will be exposed to more information about AI in mammography. This could come in the form of hospital websites, commercial entities, educational materials, and news outlets. This may lead women to search for AI-related information online and affect their medical decision-making.
However, Araen and co-authors noted a lack of data on the readability and understandability of online patient education materials for AI in mammography. To this end, the researchers identified informational barriers in communicating AI’s use in mammography to women. They searched Google for content in this area and included the following data points: the proportion of webpages that are online patient education materials and the type, readability, reading ease, and understandability of the content.
The team collected the top 20 nonsponsored results for each of five AI-related internet search terms in mammography. It included 56 webpages in its final analysis, categorizing by source type and readability by using six readability algorithms. The team also categorized by understandability using the Patient Education Materials Assessment Tool for Printable Materials.
Collectively, the webpages achieved an average grade-level readability of a college sophomore. This is far above the recommendations set by the American Medical Association (6th grade) and the Centers for Disease Control and Prevention (8th grade).
Also, reading ease scores placed most content in the “difficult” range, which is most suitable for college-level readers.
Finally, understandability averaged 72.4%, showing good understandability. However, this varied by source type. For example, pages focused on commercial, government, and patient advocacy scored the highest, while academic and medical media sources scored the lowest.
The findings could help guide clinicians in choosing the best ways of communicating this information to patients, the study authors highlighted. They called for a follow-up study to further study the extent to which readability and understandability may influence future behavior among women seeking AI mammography information.
“Developing standardized, patient-focused education materials by professional medical societies could ensure access to comprehensible information,” the authors added.
Read the full study here.
