How AI May Help Patients With Cancer and Oncology Nurses

Artificial intelligence (AI) may help improve symptom management, patient education and clinical workflows in oncology care, but experts say human connection and nursing expertise remain central to patient care.

At the 51st Annual Oncology Nursing Society (ONS) Congress, Matthew Byrne, principal product manager for generative AI applications at Mayo Clinic, discussed how AI technologies are increasingly being integrated into oncology nursing practice while emphasizing that these tools should support — not replace — clinicians and nurses.

“The work I’m involved with is never done to replace the nurse, but rather to help expedite and standardize information gathering, and then to support our decision making,” Byrne said during the presentation.

The presentation, titled “Human First, Tech Forward: Nursing in the Age of AI,” explored both the opportunities and risks associated with AI in cancer care.

How AI could support oncology care

According to Byrne’s presentation, AI tools may help support clinical decision-making, workflow efficiency and patient-centered care. Examples highlighted in the presentation included AI-assisted symptom management, clinical trial matching, personalized care plans, pharmacogenomics and risk detection tools for patients with cancer.

The presentation also discussed “ambient AI” technologies capable of converting voice and video conversations into clinical documentation, notes, orders and workflow tasks. Byrne noted that these tools may help reduce administrative burden and improve efficiency for oncology teams.

“There’s really no way that we can keep track of all of that and put it all together very, very quickly in the moment,” Byrne said regarding the growing volume of healthcare data. “That’s where AI helps us.”

Other potential applications included AI-generated patient education materials that are personalized, multilingual, conversational and designed to support patients longitudinally throughout treatment.

“I think patient-centered education is an area that could be incredibly helpful to busy nurses and to anxious and complex patients,” Byrne said. He added that AI tools may help supplement traditional educational workflows by allowing information to be repeated, paced and tailored to individual patient needs.

“I can see that AI can be used to supplement our teaching and really allow for a slowing down of that process, so it can be paced and repeated and personalized to meet each patient’s need in that patient context,” Byrne explained.

Presentation slides additionally described AI-assisted portal messaging systems that may help clinicians draft patient responses more efficiently while reducing caregiver strain and cognitive workload.

Concerns about bias, accuracy and trust

Despite the potential benefits, Byrne also highlighted several concerns surrounding AI use in healthcare, including misinformation, bias, privacy issues and inaccurate outputs resulting from incomplete or poor-quality data.

“There continues to be concern that these AI technologies make mistakes and can resurface bias,” Byrne said.

The presentation repeatedly referenced the concept of “GIGO,” or “garbage in, garbage out,” to explain how biased or incomplete data may lead to flawed clinical recommendations or patient harm. Slides from the session warned that inaccurate or delayed data could contribute to misdiagnosis, delayed care, inequities in treatment and mistrust of AI systems.

“Bias continues to be both an overt and hidden part of healthcare,” Byrne said.

The presentation also addressed concerns about online health information quality and AI literacy, citing research suggesting many online educational videos contain low levels of reliability and accuracy.

“Unfortunately, there is evidence to suggest that nurses need to build up their tech literacy,” Byrne said. He also stressed that nurses remain responsible for reviewing AI-generated outputs before they are used in patient care.

“The nurse must ensure that the content itself, the readability and accuracy are consistently meeting institutional standards, but also the patient expectation and the patient context,” Byrne explained.

Human connection remains essential

Throughout the presentation, Byrne emphasized that AI cannot replace the compassion, judgment and human connection nurses provide in oncology care. Presentation materials stressed the importance of preserving dignity, compassion, touch and presence in increasingly technology-driven care environments.

“We have to be involved in the development, the testing and the deployment of these technologies, so that we protect our practice,” Byrne said. Closing slides stated that AI cannot “replace you or your judgement,” “match your experience as a nurse” or “mimic your empathy, support and guidance.”

The session concluded that improving AI literacy among nurses and involving clinicians in AI development and implementation will be essential as these technologies continue expanding across cancer care settings.

“We are the sentinels,” Byrne said. “There are times when we need to say, ‘Nope, that doesn’t make sense here.’”

References

1. “Human First, Tech Forward: Nursing in the Age of AI” by Matthew Byrne, presented at the 51st Annual Oncology Nursing Society Congress.

For more news on cancer updates, research and education, don’t forget to subscribe to CURE®’s newsletters here.