A new study finds that artificial intelligence could encourage doctors to stop prescribing drugs that aren’t necessary — especially in older folks who tend to be on multiple medications. The report was published Thursday in the Journal of Medical Systems.
More than 40% of seniors are on five or more prescription medications — a rate that raises their risk for potentially harmful drug interactions.
Researchers used ChatGPT to assess a faux medication list for seniors. The team requested the AI technology to evaluate a variety of clinical scenarios when a patient took different mixtures of medications. The scenarios included having a history of heart disease and having a difficult time performing activities of daily living.
The program recognized drugs that weren’t necessary and recommended that doctors no longer prescribe the drug. When it came to heart disease, the platform was more likely not to suggest changes to the patient’s medication. Daily impairments didn’t seem to affect what the AI platform suggested.
“Our findings suggest that AI-based tools can play an important role in ensuring safe medication practices for older adults,” Arya Rao, a one of the researchers who is also a student at Harvard Medical School, said in a statement. “It is imperative that we continue to refine these tools to account for the complexities of medical decision-making.”
AI isn’t a perfect tool compared to having an actual doctor. For instance, it tended not to pay much attention to a person’s pain level, suggesting that the patient stop taking pain medications but not others for clinical ailments such as high cholesterol or high blood pressure, the authors reported.
“While caution should be taken to increase accuracy of such models, AI-assisted [medication] management could help alleviate the increasing burden on general practitioners,” Marc Succi, MD, a researcher and associate chair of innovation and commercialization at Mass General Brigham Radiology in Boston, said in a statement.
“Further research with specifically trained AI tools may significantly enhance the care of aging patients,” Succi said.
