Although treatment and survival rates for cancer are improving, a cancer diagnosis of any kind still can be an overwhelming scenario for older adults.
Artificial intelligence may be able to assist seniors’ caregivers in offering support and — perhaps most crucially — information about radiation oncology treatment, a new study shows.
Approximately half of all people being treated for cancer receive radiation therapy as part of their treatment regimen, according to the Mayo Clinic. Within long-term care, 10% of nursing home residents will receive a cancer diagnosis at some point during their stay, studies show.
“This is a really technical field that can be hard to understand,” study author Amulya Yalamanchili, MD, said in a statement. “If [patients] have cancer in a sensitive area, they may not feel comfortable asking what their life will look like long term. The hope is that patients can educate themselves with ChatGPT before and after they see a physician.”
The research, which was conducted by experts at Northwestern University, found that ChatGPT answered 115 common patient questions as well as or better than professionals. The responses were at a college reading level and noted that the AI missed minor details, such as positioning tattoos used as part of cancer radiation treatment in special cases.
The scientists are hoping to address certain issues and refine the responses as part of an in-house app they are currently working on for the next phase of the research project.
The use of AI in healthcare and long-term care, for cancer or any other condition, remains a contentious issue. These and other studies show that AI may be beneficial in providing older adults with useful health information, particularly if caregivers don’t know the answer.
Experts, however, have raised concerns about using AI in place of caregivers or having AI tools attempt to mimic empathy to simulate personal interaction.
In addition, although AI can be valuable as a tool for general information, it remains problematic as a diagnostic aid.
Although some AI models show a promising ability to detect early onset of various conditions, from cancers to dementia, some models continue to produce errors in contexts where the stakes are too high to effectively rely on the technology.
The Northwestern study was published earlier this month in the journal JAMA Open Network.
